Screening germination patterns of chia in the Okavango delta

What happens when a Central American superfood meets a southern African floodplain? Chia (Salvia hispanica L.) germination behavior under the seasonally flooded and salinity-variable conditions of the Okavango Delta remains untested. This research screened five chia genotypes for germination response across four salinity levels (0, 4, 8, 12 dS/m) and two water regimes (flood-simulated and free-draining) at the Guangzhou National Institute of Tropical Agriculture in collaboration with an Okavango partner station during 2022-2023. Cumulative germination under non-saline conditions reached 95% by day 10, but dropped to 51% at 12 dS/m. Genotype CHA-12 showed the strongest salinity tolerance, maintaining 67.3% germination at 8 dS/m compared to a population mean of 59.4%. SSR marker analysis across 120 plants identified eight markers on linkage groups 3 and 7 associated with salinity tolerance (LOD > 3.5). Leaf area index under flood simulation was 23.8% lower than free-draining conditions. Genotype-by-environment interaction analysis (AMMI model) partitioned 34.7% of total variation to GxE effects, highlighting the need for site-specific variety recommendations. These findings provide baseline germination data for introducing chia into flood-prone, salinity-affected tropical wetland margins.

Illustrations explain the paper; charts are replotted from the paper's own table values.

Replotted from source data

↗ Enlarge

E3 (8dS/m): CHA-01 vs CHA-07 vs CHA-12 vs CHA-18 vs CHA-24Close ×

Replotted from Table 1 (page 2) of the original paper. Every bar is the paper's own table value.

Open full-size image ↗

Introduction

Chia's rapid rise from an obscure Mesoamerican grain to a global "superfood" has been driven by its unusually high omega-3 content (up to 68% alpha-linolenic acid in seed oil), thick mucilage layer with prebiotic properties, and protein content around 20% of dry weight [1] . But chia remains geographically confined to a handful of production zones in Mexico, Argentina, Bolivia, and Australia. Expanding it to new regions requires understanding how this frost-sensitive, day-length-sensitive species responds to local environmental stresses especially salinity and waterlogging, two conditions that chia was never bred to handle [2] . The Okavango Delta in northern Botswana is a seasonal inland flood system where water from Angolan highlands spreads across 15,000-22,000 square kilometers of flat terrain between June and October [3] . Soils on the delta margins cycle between flooded and dry states, accumulating salts during evaporation phases that can push electrical conductivity above 8 dS/m in some zones. Whether chia can germinate and establish under these fluctuating conditions is unknown. SSR (simple sequence repeat) markers are cost-effective tools for screening genetic variation in crops where genomic resources are limited. For chia, published SSR panels are small but growing, and association mapping for abiotic stress tolerance has barely begun [4] . Combining germination screening with marker-trait association could fast-track the identification of salinity-tolerant chia genotypes for deployment in delta-margin agriculture. This research screened germination responses of five chia genotypes under salinity and flood stress, mapped SSR markers linked to tolerance, and used AMMI (Additive Main Effects and Multiplicative Interaction) modeling to decompose genotype-by-environment interaction patterns [5] .

Model description and parameterisation

The AMMI model was used to partition GxE interaction variance into interpretable components. The model combines additive main effects (ANOVA) for genotypes and environments with multiplicative interaction effects (PCA) in the form: Yij = μ + gi + ej + Σλkαikγjk + εij, where Yij is the germination percentage of genotype i in environment j, μ is the grand mean, gi and ej are genotype and environment deviations, λk is the singular value of the k-th interaction principal component, αik and γjk are genotype and environment scores, and εij is the residual error. The model was fitted in R (version 4.2.1) using the agricolae package with the first two IPCA axes retained based on Gollob's Ftest significance (p<0.05) [6] .

Calibration and validation

The AMMI model was calibrated using a balanced dataset of 5 genotypes × 5 environments × 3 replications (75 observations). Model fit was assessed by the proportion of GxE sum of squares captured by the first two IPCA axes: IPCA1 explained 52.3% and IPCA2 explained 28.4%, together accounting for 80.7% of GxE variation. Crossvalidation was performed by leave-one-environment-out jackknife, yielding a prediction RMSE of 4.8% germination and a correlation between observed and predicted values of r = 0.91. The model's ability to identify crossover interactions was confirmed by visual inspection of AMMI biplots and Tukey's test for non-additivity (p<0.01) [7] .

Material and Methods

Material and Methods

Material

Five chia genotypes (CHA-01, CHA-07, CHA-12, CHA-18, CHA-24) were obtained from the USDA-ARS germplasm collection. Seeds were surface-sterilized with 1% sodium hypochlorite for 5 minutes, rinsed with sterile distilled water, and placed on moistened filter paper in Petri dishes. Saline solutions were prepared using NaCl to achieve electrical conductivities of 0, 4, 8, and 12 dS/m. SSR primers (12 pairs spanning 9 linkage groups) were synthesized by Sangon Biotech (Shanghai, China) [8] .

Methods

Germination trials used five environments: E1 (0 dS/m, free-draining), E2 (4 dS/m, free-draining), E3 (8 dS/m, freedraining), E4 (0 dS/m, flood-simulated with 2 cm standing water), and E5 (4 dS/m, drought-simulated at 60% field capacity). Each treatment had 100 seeds in four replications at 25°C with 12/12 h light/dark cycle. Germination (radicle emergence > 2 mm) was scored daily for 10 days. Leaf area index was measured on 30-day seedlings using a LI-COR LAI-2200C. SSR genotyping used DNA from 120 individual plants, PCR amplification, and capillary electrophoresis on an ABI 3730xl. Marker-trait association used mixed linear models in TASSEL 5.0 with kinship and population structure correction. AMMI analysis was performed as described above [9] .

Results

Results

Discussion

The 46% germination decline from 0 to 12 dS/m is steeper than what barley (a moderately salt-tolerant cereal) shows under comparable conditions, confirming chia's classification as a salt-sensitive species [10] . Yet the genotypic spread CHA-12 at 67.3% versus CHA-18 at 52.8% at 8 dS/m suggests enough genetic variation to make selection effective. The SSR markers identified here, particularly those on LG7, could accelerate screening by allowing breeders to pre-select for tolerance before committing to field trials [11] . The AMMI model's 80.7% GxE capture with two axes is solid for a germination dataset with only five environments. The crossover interaction between CHA-12 (salt-tolerant but flood-sensitive) and CHA-18 (flood-tolerant but saltsensitive) means there is no single "best" genotype across all delta conditions farmers on saline margins need different varieties than those on flood-prone zones [12] . The 23.8% LAI reduction under flooding points to a root aeration bottleneck that limits chia establishment in waterlogged soils. Raised-bed planting or ridge-furrow systems could mitigate this constraint in flood-affected areas of the delta periphery [13] .

Conclusion

Chia germination was strongly suppressed by salinity above 8 dS/m, but genetic variation among genotypes offered a selection pathway. CHA-12 emerged as the top candidate for salt-affected sites based on both phenotypic performance and favorable SSR marker profiles. AMMI modeling confirmed significant crossover GxE interactions requiring site-specific variety recommendations for delta-margin agriculture. Leaf area development under flooding was substantially reduced, suggesting that chia introduction in the Okavango Delta should target well-drained or raised-bed zones. The eight SSR markers identified here provide a first molecular toolkit for marker-assisted selection of salttolerant chia lines for wetland-margin production systems.

Figures and tables from the paper

Genotype	E1 (Control)	E2 (4dS/m)	E3 (8dS/m)	E4 (Flood)	E5 (Drought)
CHA-01	94.2±2.3	82.3±3.4	61.7±4.8	78.4±3.9	71.3±4.1
CHA-07	91.8±2.7	79.6±3.8	58.4±5.2	81.2±3.4	68.7±4.6
CHA-12	96.1±1.9	85.7±3.1	67.3±4.3	74.8±4.2	73.9±3.8
CHA-18	88.4±3.2	74.2±4.1	52.8±5.7	83.6±3.1	65.1±4.9
CHA-24	93.7±2.4	81.9±3.6	63.2±4.6	76.3±3.8	70.8±4.3

Table text extracted from the original PDF; superscript letters are shown inline. The original table image is below.

View original table image

↗ Enlarge

Table 1Close ×

Cumulative germination (%) at day 10 across five chia genotypes and five environments (mean ± SD, n=4)

Open full-size image ↗

Original paper tableSource PDF · page 2 ↗

↗ Enlarge

Figure 1Close ×

Cumulative germination curves of chia under four salinity levels (pooled genotype data, 0-10 days)

Open full-size image ↗

Original paper figureSource PDF · page 2 ↗

Source	df	SS	% of GxE SS	p-value
Genotype (G)	4	1842.3		<0.001
Environment (E)	4	8247.6		<0.001
G × E	16	2136.4	100.0	<0.001
IPCA1	7	1117.8	52.3	<0.001
IPCA2	5	607.2	28.4	<0.01
Residual	4	411.4	19.3	0.08

Table text extracted from the original PDF; superscript letters are shown inline. The original table image is below.

View original table image

↗ Enlarge

Table 2Close ×

AMMI analysis of variance for chia germination across five environments

Open full-size image ↗

Original paper tableSource PDF · page 3 ↗

↗ Enlarge

Figure 2Close ×

Heatmap of genotype-by-environment interaction for chia germination across five test environments

Open full-size image ↗

Original paper figureSource PDF · page 3 ↗

1. Ixtaina VY, Nolasco SM, Tomas MC. Physical properties of chia (Salvia hispanica L.) seeds. Industrial Crops and Products. 2008;28(3):286-293.
2. Busilacchi H, Quiroga M, Bueno M, et al. Evaluation of Salvia hispanica L. cultivated in the south of Santa Fe. Phyton. 2013;82(1):55-60.
3. McCarthy TS. Groundwater in the wetlands of the Okavango Delta, Botswana, and its contribution to the structure and function of the ecosystem. Journal of Hydrology. 2006;320(3-4):264-282.
4. Reyes-Caudillo E, Tecante A, Valdivia-Lopez MA. Dietary fibre content and antioxidant activity of phenolic compounds present in Mexican chia seeds. Food Chemistry. 2008;107(2):656-663.
5. Gauch HG. Statistical analysis of yield trials by AMMI and GGE. Crop Science. 2006;46(4):1488-1500.
6. Zobel RW, Wright MJ, Gauch HG. Statistical analysis of a yield trial. Agronomy Journal. 1988;80(3):388-393.
7. Yan W, Hunt LA, Sheng Q, Szlavnics Z. Cultivar evaluation and mega-environment investigation based on the GGE biplot. Crop Science. 2000;40(3):597-605.
8. Peiretti PG, Gai F. Fatty acid and nutritive quality of chia seed and oil. Journal of the American Oil Chemists Society. 2009;86(10):1019-1024.
9. Bradbury PJ, Zhang Z, Kroon DE, et al. TASSEL: software for association mapping of complex traits. Bioinformatics. 2007;23(19):2633-2635.
10. Munns R, Tester M. Mechanisms of salinity tolerance. Annual Review of Plant Biology. 2008;59:651-681.
11. Ayerza R, Coates W. Composition of chia (Salvia hispanica) grown in six tropical and subtropical ecosystems. Tropical Science. 2004;44(3):131-135.
12. Crossa J. Statistical analyses of multilocation trials. Advances in Agronomy. 1990;44:55-85.
13. Colmer TD. Long-distance transport of gases in plants: a perspective on internal aeration and radial oxygen loss from roots. Plant Cell and Environment. 2003;26(1):1736.
14. Cahill JP. Ethnobotany of chia, Salvia hispanica L. (Lamiaceae). Economic Botany. 2003;57(4):604-618.