White Paper-Figures-PB-KJB.pdf

Figure 1. Examples of oxygen consumption curves and illustration of the derivation of the ASTEC values as well as R50 and AUC. Typical curves are shown for a rapidly germinating seed (left curve) and a slowly germinating seed (right curve). Seeds generally exhibit the two-stage or sigmoidal patterns shown here, but some seeds have linear patterns or other variations. Brief descriptions of the derived ASTEC values and other indices are listed in the figure; see text for further explanation. The AUC50 for the slowly germinating seed is not shaded for clarity. From Bradford (2013).

Figure 2. Example of the generation of population oxygen depletion (POD) curves for tomato seeds imbibed at 25°C. Initial oxygen percentage is normalized to 100% (equivalent to 21% air composition) and each grey curve indicates the oxygen consumption time course of an individual seed. The colored lines indicate the cumulative percentages of individual seeds over time that have reduced the oxygen contents of sealed wells (or vials) to 75% (R75 – solid blue line), 50% (R50 – solid green line) and 25% (R25 – solid orange line) of the original level. The colored circles on the time axis indicate the median times required by the population of individual seeds to reduce oxygen levels to 75, 50 and 25% of the initial value (R75(50), R50(50) and R25(50) values, respectively). Time courses can also be constructed based on other oxygen consumption percentages. Black lines represent the median (solid) and average (dashed) oxygen depletion time courses for the entire seed population. From Bradford and Bello (2016).

Cu m ul at ive se ed pe rc en ta ge

R25

R50

R75

0 20 40 60 80 100 120 140

Time (h)

O xy ge n of in itia l(

Figure 3. Oxygen consumption time courses of individual untreated (control) radish seeds (A) that had been subjected to 45 (B), 66 (C), 80 (D) days of controlled deterioration at 33% RH and 50°C. Similar curves are shown for individual lettuce seeds that were untreated (E) or had been subjected to 2 (F), 4 (G), or 6 (H) days of controlled deterioration at 75% RH and 50°C. Control lettuce seeds were also subjected to priming (prehydration and drying) prior to measurement (I). Each curve indicates the oxygen consumption time course of an individual seed, and is color coded to reflect the time by which radicle emergence had occurred for that seed (see legends in panels A, E and I and also the colored bars on the x-axes). Additional grey curves represent non-germinating seeds. Averages (dashed curves) and medians

(solid lines) for the seed populations are shown in black. From Bradford and Bello (2016).

10 20 30 40 50 60 70 Time (h)

12h 16h 20h 24h 32h

Primed I

Germ. Time

Control E

Aged 2 days F

Aged 4 days G

0 10 20 30 40 50 60 70 Time (h)

12h 16h 20h 24h 32h 40h 48h 72h

Aged 6 days H

O xy ge n of In itia l)

Aged 45 days B

O xy ge n of In itia l)

12h 18h 24h 36h 44h 50h 60h >72h

Control A

Germ. Time

O xy ge n of In itia l)

Aged 66 days C

0 10 20 30 40 50 60

Time (h)

O xy ge n of In itia l)

Aged 80 days D

Accelerated Aging - LettuceControlled Deterioration - Radish Priming - Lettuce

Germ. Time

Figure 4. Median oxygen depletion curves of: (A) tomato seeds imbibed at temperatures of 25°C (orange), 20°C (green) or 15°C (blue) at 0 MPa; (B) tomato seeds imbibed at water potentials of 0 MPa (orange), −0.2 MPa (green), −0.4 MPa (blue), −1.0 MPa (purple), −1.5 MPa (red) or −2.0 MPa (brown) at 25°C; (C) tomato seeds imbibed in concentrations of the respiratory inhibitor KCN of 0 mM (control; orange), 0.1 mM (green), 0.2 mM (blue) or 0.3 mM (red) at 25°C; (D) tomato seeds imbibed in concentrations of the respiratory inhibitor SHAM of 0 mM (control; orange), 0.05 mM (green), 1 mM (blue), 2 mM (purple) or 3 mM (red) at 25°C. All seeds in the test populations were used to calculate median curves, regardless of their germination status. From Bradford and Bello (2016).

O xy ge n of in itia l)

25oC 20oC 15oC

TemperatureA

0 50 100 150 200

O xy ge n of in itia l)

Control -0.2 MPa -0.4 MPa -1.0 MPa -1.5 MPa -2.0 MPa

Water PotentialB

O xy ge n of in itia l)

Control KCN 0.1 mM KCN 0.2 mM KCN 0.3 mM

Respiratory Inhibitor - KCNC

0 50 100 150 200 250

Time (h)

O xy ge n of in itia l)

Control SHAM 0.5 mM SHAM 1 mM SHAM 2 mM SHAM 3 mM

Respiratory Inhibitor - SHAMD

Figure 5. Germination time courses (A, E, I and M) and R75, R50 and R25 population oxygen depletion (POD) time courses (B–D, F–H, J–L and N–P) of tomato seeds imbibed under different conditions.

Cumulative germination (A) and POD time courses (B, C, D) are shown for seeds imbibed on agar at 25°C (orange circles), 20°C (green triangles) or 15°C (blue squares). Cumulative germination (E) and POD time courses (F, G, H) are shown for seeds imbibed at 25°C on agar at 0 MPa (open orange circles), −0.2 MPa (open green triangles), −0.4 MPa (open blue squares), −1 MPa (filled purple circles), −1.5 MPa (filled red circles) or −2 MPa (filled brown circles). Cumulative germination (I) and POD time courses (J, K, L) of seeds imbibed at 25°C on agar containing 0 (control, open orange circles), 0.1 mM (open green triangles), 0.2 mM (open blue squares) or 0.3 mM (filled purple circles) KCN. Cumulative germination

(M) and POD time courses (N, O, P) of seeds imbibed at 25°C on agar containing 0 (control, open orange circles), 0.5 mM (open green triangles), 1 mM (open blue squares), 2 mM (filled purple circles) or 3 mM (filled red triangles) SHAM. Continuous curves are predicted by fitting the appropriate population-based threshold model to the data across the temperature, water potential, KCN or SHAM ranges. Solid and dashed curves indicate where two different models were fit to sub-components of the data. From Bradford and Bello (2016).

Figure 6. Relationships between median times to radicle emergence (t50) and the median times for seeds to reduce the oxygen in their wells to 75% (blue squares), 50% (green triangles) or 25% (orange circles) of the initial value for: (A) tomato seeds imbibed at different temperatures (open symbols) or water potentials (closed symbols); (B) tomato seeds imbibed in different concentrations of KCN, SHAM or KCN + SHAM (closed symbols), of ABA (open symbols) or of GA (half-closed symbols) (ABA and GA data not included in linear regressions); (C) lettuce seeds subjected to accelerated ageing (open symbols represent primed seeds, not included in the linear regressions); and (D) radish seeds subjected to controlled deterioration. Significance levels of the regression values are indicated as: ns, non-significant;

**, P < 0.01; ***, P < 0.001.

Figure 7. Germination time courses (A) and R75, R50 and R25 POD time courses (B–D) for untreated (control) radish seeds (open orange circles) or after 45 (green triangles), 66 (blue boxes) or 80 (dark blue circles) days of controlled deterioration at 33% RH and 50°C. Similarly, germination time courses (E and I) and R75, R50 and R25 POD time courses (F–H and J–L) for untreated (control) lettuce seeds (open orange circles), primed seeds (open red triangles) or seeds after 2 (green triangles), 4 (blue boxes) or 6 (purple circles) days of controlled deterioration at 75% RH and 50°C. The time courses predicted by the aging time model for germination or respiration data are represented for all treatments on panels A–H as solid lines of the corresponding color. Primed seeds were modelled using −1 day of ‘aging’ time.

Additionally, solid lines in panels I–L illustrate predicted time courses based upon summing the contributions of two distinct subpopulations in the seed lot. From Bradford and Bello (2016).

R50K

Primed Control 2 days Aged 4 days Aged 6 days Aged

R75J

0 10 20 30 40 50 60 70 Time (h)

R25L

GerminationI

Se ed s de pl et in g to O ) R50C

Se ed s de pl et in g to O

Control 45 days Aged 66 days Aged 80 days Aged

R75B

0 10 20 30 40 50 60 70

Time (h)

Se ed s de pl et in g to O ) R25D

G er m in at io n

GerminationA

Accelerated Aging - LettuceControlled Deterioration - Radish GerminationE

Primed Control 2 days Aged 4 days Aged 6 days Aged

R75F

R50G

0 10 20 30 40 50 60 70 Time (h)

R25H

Accelerated Aging - Lettuce - 2 Pop.

Figure 8. Development of subpopulation models for primed and aged lettuce seeds. The data (gray circles in panels A-D) exhibited clear evidence of subpopulations that were responding differently to priming and aging (data from Fig. 7E-H). Assuming that the seed lot was composed of two distinct subpopulations, we calculated two different models and respective parameters for different fractions of the population (panels A-D, purple and green dashed lines). Overall values obtained by summing the predicted time courses of the two subpopulations closely matched the overall respiration time courses (panels A-D, solid red lines; see also Fig. 7I-L). The normal distributions of the two subpopulations are shown based on the medians and standard deviations of the aging model (panels E-H; Table S2) that provided the best fit to the data (data for primed seeds were not used in fitting the model) (panels I-L). The summed distribution of the entire population is also shown (panels E-H, solid red lines). Subpopulation models were optimized using Excel Solver to minimize the sum of squared residuals (SSR) between actual germination/respiration and predicted values (panels I-L, red lines, solid black lines and black circles, respectively) by solving for all model parameters based on the following equation:

T(gr) = f1 {[p – (θage1 / tgr1) – pmax(50)1] / σpmax1} + (1 - f1) {[p – (θage2 / tgr2) – pmax(50)2] / σpmax2}

0 5 10 15 20 25 30 pmax(g) (days)

Pe rc en ti n se ed po pu la tio n

R25H

Pe rc en ti n se ed po pu la tio n

Whole Pop.

Subpop. 1 Subpop. 2

GerminationE

Pe rc en ti n se ed po pu la tio n

R75F

Pe rc en ti n se ed po pu la tio n

R50G

Pr ed ict ed

GerminationI y = 1.013x + 0.171 r2=0.986***

Pr ed ict ed

R75J y = 0.991x + 0.006 r2=0.989***

Pr ed ict ed

R50K y = 0.994x + 0.003 r2=0.993***

0 20 40 60 80 100

Actual (%)

Pr ed ict ed

R25L y = 0.993x + 0.003 r2=0.995***

Se ed s de pl et in g to O ) R50C

Se ed s de pl et in g to O ) R75B

G er m in at io n

Actual Data Whole Pop.

Subpop. 1 Subpop. 2

GerminationA

0 10 20 30 40 50 60 70

Time (h)

Se ed s de pl et in g to O ) R25D where T(gr) is the distribution of germination or respiration (e.g., R75, R50 or R25) times of the total population, (f1) is the fraction of the total population in subpopulation 1, p is the aging period, θage1 and θage2 are the aging time constants for subpopulations 1 and 2, tgr1 and tgr2 are the germination or respiration times of individual seeds in subpopulations 1 and 2, pmax(50)1 and pmax(50)2 are the medians of the distributions of pmax values for subpopulations 1 and 2, and σpmax1 and σpmax2 are the standard deviations of pmax values among seeds in subpopulations 1 and 2. Thus, the value of f1 and the parameters of both subpopulation models were optimized by minimizing the deviation of the total population model from the actual data (minimizing SSR; panels I-L). The curves in panels A-D illustrate the two populations by superimposing their predicted time courses on the same axis and having each population reach a maximum at its fraction of the total population (dashed blue and green lines). The figures illustrate that at longer aging times, the RXX times of seeds in the two populations overlap, resulting in loss of the distinct double-sigmoid shape that is evident in primed and control seeds (e.g., panel C). Although data for primed seeds were not included in calculations of the subpopulation equation described above, using the same model parameters with a “negative ageing” period (p = -1 day) and increasing the fraction of seeds in the faster subpopulation (f1) enabled the model to match the data well. Interestingly, even though germination was observed too infrequently to detect subpopulations directly in the data, applying the same subpopulation approach as was used for the POD curves also resulted in an improved fit of the model to the germination time course data (panels A, I). Theoretically, additional subpopulation terms could be added to the equation above (i.e., f3 = 1 - f1 - f2, etc.) if justified by the data. From Bradford and Bello (2016).

Figure 9. Q2 oxygen depletion curves of individual seeds and average and median curves for ABA-deficient sitw seeds plus ABA (A-E) or GA-deficient gib1 seeds plus GA (F-J). The time courses are colored to reflect the times by which radicle emergence had occurred for that seed (see legend in panel F). As germination was delayed (ABA) or advanced (GA) by increasing concentrations of the hormones, the respiratory patterns were affected similarly. Only seeds that transitioned to a more rapid oxygen consumption rate germinated, and the time of radicle emergence was closely associated with the time of that transition. From Bradford and Bello (2016).