Variation in seed longevity among diverse Indica rice varieties

Abstract

Background and aims: Understanding variation in seed longevity, especially within closely related germplasm, will lead to better understanding of the molecular basis of this trait, which is particularly important for seed genebanks, but is also relevant to anyone handling seeds. We therefore set out to determine the relative seed longevity of diverse Indica rice accessions through storage experiments. Since antioxidants are purported to play a role in seed storability, the antioxidant activity and phenolic content of caryopses were determined.

Methods: Seeds of 299 Indica rice accessions harvested at 31, 38 and 45 d after heading (DAH) between March and May 2015 and differing in harvest moisture content (MC) were subsequently stored at 10.9 % MC and 45 °C. Samples were taken at regular intervals and sown for germination. Germination data were subjected to probit analysis and the resulting parameters that describe the loss of viability during storage were used for genome-wide association (GWA) analysis.

Key results: The seed longevity parameters, Ki [initial viability in normal equivalent deviates (NED)], -σ-1 (σ is the time for viability to fall by 1 NED in experimental storage) and p50 [time for viability to fall to 50 % (0 NED)], varied considerably across the 299 Indica accessions. Seed longevity tended to increase as harvest MC decreased and to decrease as harvest MC increased. Eight major loci associated with seed longevity parameters were identified through GWA analysis. The favourable haplotypes on chromosomes 1, 3, 4, 9 and 11 enhanced p50 by ratios of 0.22-1.86.

Conclusions: This is the first study to describe the extent of variation in σ within a species' variety group. A priori candidate genes selected based on rice genome annotation and gene network ontology databases suggested that the mechanisms conferring high seed longevity might be related to DNA repair and transcription, sugar metabolism, reactive oxygen species scavenging and embryonic/root development.

Keywords: Oryza sativa L; GWAS; Genebank; Indica variety group; rice; seed longevity; seed storage; viability monitoring.

Fig. 1.

Changes in seed harvest moisture content (MC) and 100-seed weight over the harvesting period for 299 Indica rice accessions. Seeds were harvested at 31, 38 and 45 days after heading (DAH). Panels (B) and (D) are box plots of the data shown in (A) and (C), to summarize the changes in harvest MC and 100-seed weight for the different harvests. In (B) and (D) the square symbol is the mean, the box spans the 25th and 75th percentiles, the whiskers represent the 5th and 95th percentiles, the crosses the 1st and 99th percentiles, and the dashes the minimum and maximum values, respectively.

Fig. 2.

Relationships between the change in longevity (time for viability to fall to 50 % during experimental storage, p₅₀) and change in harvest MC for Indica rice seeds that either (A) lost moisture between sequential harvests [i.e. between 31 and 38 days after heading (DAH) or between 38 and 45 DAH] or (B) gained moisture. The change in p₅₀ was calculated as a proportion of the p₅₀ of the first harvest (i.e. 31 or 38 DAH, respectively). The red lines show the results of linear regression analysis. Seeds were stored at 45 °C and 10.77–11.66 % moisture content.

Fig. 3.

Three-dimensional plots showing the relationship between K_i, –σ⁻¹ and p₅₀ (coloured surface) and the estimates of these parameters for seeds of 299 Indica rice accessions (A) harvested at 38 days after heading (DAH) or (B) with the highest estimate of p₅₀ (harvested at 38 or 45 DAH). Seeds were stored at 45 °C and 10.77–11.66 % moisture content.

Fig. 4.

Correlations between 100-seed weight, p_50(38DAH), p_50(max), caryopsis antioxidant activity and caryopsis phenolic content for 299 Indica rice accessions. Correlation coefficients (r) were significant at *P < 0.05 or ***P < 0.001 (two-sided tests).

Fig. 5.

GWA analysis of K_i, –σ⁻¹ (slope) and p₅₀ for seeds harvested at 38 days after heading (DAH). Harvest moisture content was included as a covariate of K_i and p₅₀. Plots above the red threshold line are significantly associated with traits. Different colours refer to the 12 different chromosomes (see box).

Fig. 6.

GWA analysis of p₅₀ of Indica rice accessions harvested at (A) 38 days after heading (DAH) and (B) either 38 or 45 DAH (whichever gave the highest p₅₀ for each accession). Harvest moisture content was used as a covariate of p₅₀. Plots above the red threshold line are significantly associated with the trait. Different colours refer to the 12 different chromosomes (see box). Arrows indicate loci that appeared when the maximum p₅₀ for each accession was used.

Fig. 7.

Haplotype effects on seed longevity (p₅₀) in a diverse Indica rice panel.