Environmental context of phenotypic plasticity in flowering time in sorghum and rice

Abstract Phenotypic plasticity is an important topic in biology and evolution. However, how to generate broadly applicable insights from individual studies remains a challenge. Here, with flowering time observed from a large geographical region for sorghum and rice genetic populations, we examine the consistency of parameter estimation for reaction norms of genotypes across different subsets of environments and searched for potential strategies to inform the study design. Both sample size and environmental mean range of the subset affected the consistency. The subset with either a large range of environmental mean or a large sample size resulted in genetic parameters consistent with the overall pattern. Furthermore, high accuracy through genomic prediction was obtained for reaction norm parameters of untested genotypes using models built from tested genotypes under the subsets of environments with either a large range or a large sample size. With 1428 and 1674 simulated settings, our analyses suggested that the distribution of environmental index values of a site should be considered in designing experiments. Overall, we showed that environmental context was critical, and considerations should be given to better cover the intended range of the environmental variable. Our findings have implications for the genetic architecture of complex traits, plant–environment interaction, and climate adaptation.

*Corresponding authors Tingting Guo, tguo@mail.hzau.edu.cn; and Jianming Yu, jmyu@iastate.edu The following Supplementary Data is available for this article: Supplementary Figure S1 -S12 Supplementary Table S1 -      Table S1.Description of the nine environments used for multi-environment trials in sorghum and rice.Table S2.Flowering time measured as the accumulated growing degree days in nine environments each having two replications in sorghum (separate file).Table S3.Flowering time measured as the days after planting in nine environments each having two replications in rice (separate file).Table S4.Genotypic information of 1426 SNPs for 237 recombinant inbred lines (separate file).Table S5.Genotypic information of 162 restriction fragment length polymorphic markers for 176 backcross inbred lines (separate file).Table S6.Weather data in the empirical multi-environment trials in sorghum (separate file).Table S7.Weather data in the empirical multi-environment trials in rice (separate file).Table S8.Weather data in the simulation study in sorghum (separate file).Table S9.Weather data in the simulation study in rice (separate file).Table S10.Description of empirical and simulated experiments across years and planting dates.Table S11.Variance partitioning of environmental index into site, year, planting date, and residuals in the simulated experiments.Table S12.Variance partitioning of environmental index within each testing site into year, planting date, and residuals.Table S2.Flowering time measured as the accumulated growing degree days in nine environments each having two replications in sorghum (separate file).Table S3.Flowering time measured as the days after planting in nine environments each having two replications in rice (separate file).Table S4.Genotypic information of 1426 SNPs for 237 recombinant inbred lines (separate file).Table S5.Genotypic information of 162 restriction fragment length polymorphic markers for 176 backcross inbred lines (separate file).Table S6.Weather data in the empirical multi-environment trials in sorghum (separate file).Table S7.Weather data in the empirical multi-environment trials in rice (separate file).Table S8.Weather data in the simulation study in sorghum (separate file).Table S9.Weather data in the simulation study in rice (separate file).

S12Figure S1 .
Figure S1.The design of this study to investigate the consistency of phenotypic plasticity estimation.Figure S2.Genomic prediction of phenotypic plasticity.Figure S3.Sorghum planting dates in Kansas across 12 years.Figure S4.Reaction norms of subsets of environments with small, medium, and large environmental index ranges.Figure S5.Correlations between the whole set and subsets of two environments for slope and intercept in sorghum.Figure S6.Correlations between the whole set and subsets of two environments for slope and intercept in rice.Figure S7.Effects of environment sample size and environmental mean range on correlations of flowering time phenotypic plasticity between subsets and the whole set.Figure S8.Fitted lines of slope estimations and predictions for recommending environment sample size and environmental mean range.Figure S9.Fitted lines of intercept estimations and predictions for recommending environment sample size and environmental mean range.Figure S10.Comparison of phenotypic plasticity estimates between the whole set of nine environments and two contrasting environments in rice.Figure S11.Changes in environmental index values across years and planting dates in testing sites in Iowa, Kansas, and Puerto Rico.Figure S12.Changes in environmental index values across years and planting dates in TS, FU, ISA, ISI, TH, and HA.

Figure S3 .
Figure S1.The design of this study to investigate the consistency of phenotypic plasticity estimation.Figure S2.Genomic prediction of phenotypic plasticity.Figure S3.Sorghum planting dates in Kansas across 12 years.Figure S4.Reaction norms of subsets of environments with small, medium, and large environmental index ranges.Figure S5.Correlations between the whole set and subsets of two environments for slope and intercept in sorghum.Figure S6.Correlations between the whole set and subsets of two environments for slope and intercept in rice.Figure S7.Effects of environment sample size and environmental mean range on correlations of flowering time phenotypic plasticity between subsets and the whole set.Figure S8.Fitted lines of slope estimations and predictions for recommending environment sample size and environmental mean range.Figure S9.Fitted lines of intercept estimations and predictions for recommending environment sample size and environmental mean range.Figure S10.Comparison of phenotypic plasticity estimates between the whole set of nine environments and two contrasting environments in rice.Figure S11.Changes in environmental index values across years and planting dates in testing sites in Iowa, Kansas, and Puerto Rico.Figure S12.Changes in environmental index values across years and planting dates in TS, FU, ISA, ISI, TH, and HA.

Figure S5 .
Figure S1.The design of this study to investigate the consistency of phenotypic plasticity estimation.Figure S2.Genomic prediction of phenotypic plasticity.Figure S3.Sorghum planting dates in Kansas across 12 years.Figure S4.Reaction norms of subsets of environments with small, medium, and large environmental index ranges.Figure S5.Correlations between the whole set and subsets of two environments for slope and intercept in sorghum.Figure S6.Correlations between the whole set and subsets of two environments for slope and intercept in rice.Figure S7.Effects of environment sample size and environmental mean range on correlations of flowering time phenotypic plasticity between subsets and the whole set.Figure S8.Fitted lines of slope estimations and predictions for recommending environment sample size and environmental mean range.Figure S9.Fitted lines of intercept estimations and predictions for recommending environment sample size and environmental mean range.Figure S10.Comparison of phenotypic plasticity estimates between the whole set of nine environments and two contrasting environments in rice.Figure S11.Changes in environmental index values across years and planting dates in testing sites in Iowa, Kansas, and Puerto Rico.Figure S12.Changes in environmental index values across years and planting dates in TS, FU, ISA, ISI, TH, and HA.

Figure S6 .
Figure S1.The design of this study to investigate the consistency of phenotypic plasticity estimation.Figure S2.Genomic prediction of phenotypic plasticity.Figure S3.Sorghum planting dates in Kansas across 12 years.Figure S4.Reaction norms of subsets of environments with small, medium, and large environmental index ranges.Figure S5.Correlations between the whole set and subsets of two environments for slope and intercept in sorghum.Figure S6.Correlations between the whole set and subsets of two environments for slope and intercept in rice.Figure S7.Effects of environment sample size and environmental mean range on correlations of flowering time phenotypic plasticity between subsets and the whole set.Figure S8.Fitted lines of slope estimations and predictions for recommending environment sample size and environmental mean range.Figure S9.Fitted lines of intercept estimations and predictions for recommending environment sample size and environmental mean range.Figure S10.Comparison of phenotypic plasticity estimates between the whole set of nine environments and two contrasting environments in rice.Figure S11.Changes in environmental index values across years and planting dates in testing sites in Iowa, Kansas, and Puerto Rico.Figure S12.Changes in environmental index values across years and planting dates in TS, FU, ISA, ISI, TH, and HA.

Figure S11 .
Figure S1.The design of this study to investigate the consistency of phenotypic plasticity estimation.Figure S2.Genomic prediction of phenotypic plasticity.Figure S3.Sorghum planting dates in Kansas across 12 years.Figure S4.Reaction norms of subsets of environments with small, medium, and large environmental index ranges.Figure S5.Correlations between the whole set and subsets of two environments for slope and intercept in sorghum.Figure S6.Correlations between the whole set and subsets of two environments for slope and intercept in rice.Figure S7.Effects of environment sample size and environmental mean range on correlations of flowering time phenotypic plasticity between subsets and the whole set.Figure S8.Fitted lines of slope estimations and predictions for recommending environment sample size and environmental mean range.Figure S9.Fitted lines of intercept estimations and predictions for recommending environment sample size and environmental mean range.Figure S10.Comparison of phenotypic plasticity estimates between the whole set of nine environments and two contrasting environments in rice.Figure S11.Changes in environmental index values across years and planting dates in testing sites in Iowa, Kansas, and Puerto Rico.Figure S12.Changes in environmental index values across years and planting dates in TS, FU, ISA, ISI, TH, and HA.

Figure S12 .
Figure S1.The design of this study to investigate the consistency of phenotypic plasticity estimation.Figure S2.Genomic prediction of phenotypic plasticity.Figure S3.Sorghum planting dates in Kansas across 12 years.Figure S4.Reaction norms of subsets of environments with small, medium, and large environmental index ranges.Figure S5.Correlations between the whole set and subsets of two environments for slope and intercept in sorghum.Figure S6.Correlations between the whole set and subsets of two environments for slope and intercept in rice.Figure S7.Effects of environment sample size and environmental mean range on correlations of flowering time phenotypic plasticity between subsets and the whole set.Figure S8.Fitted lines of slope estimations and predictions for recommending environment sample size and environmental mean range.Figure S9.Fitted lines of intercept estimations and predictions for recommending environment sample size and environmental mean range.Figure S10.Comparison of phenotypic plasticity estimates between the whole set of nine environments and two contrasting environments in rice.Figure S11.Changes in environmental index values across years and planting dates in testing sites in Iowa, Kansas, and Puerto Rico.Figure S12.Changes in environmental index values across years and planting dates in TS, FU, ISA, ISI, TH, and HA.

Figure S1 .
Figure S1.The design of this study to investigate the consistency of phenotypic plasticity estimation.Phenotypic plasticity parameters (intercept and slope) are obtained from the original trait values.The estimated parameter values are compared between subsets and the whole set of environments using correlation and prediction methods.These comparisons provide guidance for selection and recommendation of an optimal set of environments for studying phenotypic plasticity.

Figure S2 .
Figure S2.Genomic prediction of phenotypic plasticity.(A) Scenario A: genomic prediction of phenotypic plasticity is performed within subsets of environments.(B) Scenario B: genomic prediction of phenotypic plasticity from subsets of environments is compared with estimates obtained from the whole set.

Figure S3 .
Figure S3.Sorghum planting dates in Kansas across 12 years.On average, about 30% fields are planted in the 22 nd week and almost all fields are planted in the 26 th week of the year (Data from https://www.nass.usda.gov/)(Using Week as the x-axis).

Figure S4 .
Figure S4.Reaction norms of subsets of environments with small, medium, and large environmental mean ranges.(A-D) Reaction norms of subsets with different ranges and the whole set of environments in sorghum.(E-H) Reaction norms of subsets with different ranges and the whole set of environments in rice.

Figure S5 .
Figure S5.Correlations between the whole set and subsets of two environments for slope and intercept in sorghum.The subsets of two environments with small (A-B), medium (C-D), and large (E-F) environmental mean ranges result in different correlations of slope and intercept.

Figure S6 .
Figure S6.Correlations between the whole set and subsets of two environments for slope and intercept in rice.The subsets of two environments with small (A-B), medium (C-D), and large (E-F) environmental mean ranges result in different correlations of slope and intercept.

Figure S7 .
Figure S7.Effects of environment sample size and environmental mean range on correlations of flowering time phenotypic plasticity between subsets and the whole set.The subsets of environments are categorized into three groups: two environments or 2E, three environments or 3E, and more than three environments (>3E).(A, C) Correlations between subsets and the whole set for slopes in sorghum (A) and rice (C).(B, D) Correlations between subsets and the whole set for intercepts in sorghum (B) and rice (D).

Figure S8 .
Figure S8.Fitted lines of slope estimations and predictions for recommending environment sample size and environmental mean range.(A, D) Correlations between subsets and the whole set of environments for slopes in sorghum (A) and rice (D).(B, E) Genomic prediction accuracy for slopes of untested genotypes within subsets of environments in sorghum (B) and rice (E) (Scenario A). (C, F) Genomic prediction accuracy for slopes of untested genotypes across the whole range in sorghum (C) and rice (F) (Scenario B).Scenario A and Scenario B are shown in Figure S2.The number in the parentheses represents how well the regression model fits the data.

Figure S9 .
Figure S9.Fitted lines of intercept estimations and predictions for recommending environment sample size and environmental mean range.(A, D) Correlations between subsets and the whole set of environments for intercepts in sorghum (A) and rice (D).(B, E) Genomic prediction accuracy for intercepts of untested genotypes within subsets of environments in sorghum (B) and rice (E) (Scenario A). (C, F) Genomic prediction accuracy for intercepts of untested genotypes across the whole range in sorghum (C) and rice (F) (Scenario B).Scenario A and Scenario B are shown in Figure S2.The number in the parentheses represents how well the regression model fits the data.

Figure S10 .
Figure S10.Comparison of phenotypic plasticity estimates between the whole set of nine environments and two contrasting environments in rice.(A) Reaction norms across nine environments.(B) Reaction norms from two environments with the largest environmental mean range.(C-D) Correlation between nine and two environments for slope (C) and intercept (D).

Figure S11 .
Figure S11.Changes in environmental index values across years and planting dates in testing sites in Iowa, Kansas, and Puerto Rico.(A-B) Variability of the environmental index PTT18-43 (photothermal time from 18-43 days after planting) observed in Iowa (A) and Kansas (B) across different years and planting dates.(C-D) Consistency of the environmental index PTT18-43 observed in Puerto Rico during both summer (C) and winter (D), regardless of the year or planting date.

Figure S12 .
Figure S12.Changes in environmental index values across years and planting dates in TS, FU, ISA, ISI, TH, and HA.(A-B) Variability of the environmental index GDD9-50 (Growing degree days from 9-50 days after planting) observed in TS&FU and ISA, with greater variation across planting dates but less variation across years.(C) The environmental index GDD9-50 in ISI shows an increase with delayed plantings but remains consistent across years.(D) The environmental index GDD9-50 shows slight variation among years and planting dates in TH&HA.TS and FU sites are combined, as well as TH and HA sites, due to overlapping distributions and close coordinates.

Table S10 .
Description of empirical and simulated experiments across years and planting dates.

Table S11 .
Variance partitioning of environmental index into site, year, planting date, and residuals in the simulated experiments.

Table S12 .
Variance partitioning of environmental index within each testing site into year, planting date, and residuals.