Professional Documents
Culture Documents
Table 1. Site information for collections of Bromus tectorum seeds used in experiments (adapted from Meyer et al., 1997)
Collection site Ecological type Elevation Plant community Mean Mean Mean
( USA) (m) type annual temperature temperature
precipitation January July
(cm) (°C ) (°C )
−1.27
−1.15
−1.20
−0.81
25 °C
−1.15
−1.20
−1.11
−1.27
15 °C
b
20/30 °C
+0.50
+0.84
−0.16
+0.14
10/20 °C
−0.13
−0.12
+0.12
−0.81
y (50) [recently harvested ]
+0.77
+0.17
+0.86
25 °C
regression analysis to obtain the base temperature with the best fit, the
b
dotted horizontal line is the y (50) value for fully after-ripened seeds,
–
b
b
the slope of the regression line gives the decrement in y (50) per unit
b
of thermal time, and the intercept of the regression line with the dotted
horizontal line gives h , the thermal time requirement for afterripening.
yb
AR
See Table 3 for parameter values.
0.79
0.90
0.83
0.82
R2
43
35
22
52
Strawberry
Discussion
Several decisions were made during the course of model
building to meet the challenges that were encountered.
Many of these decisions required assumptions, whether
theoretical or necessary from a practical standpoint. The
ability of the model successfully to predict rates of dorm-
ancy loss in the field suggests that the use of these
assumptions was appropriate.
Fig. 3. Predicted and observed changes in y (50) during field after- This model is based on the previously-verified assump-
b tion (Christensen et al., 1996) that differences in germina-
ripening for seeds from four populations of Bromus tectorum as
measured at two incubation temperatures. Symbols represent measured tion time-course curves of a seed population due to after-
laboratory values for seeds retrieved from the field, solid lines represent
values predicted from simulation modelling, and dotted horizontal lines ripening and incubation temperature can be described by
represent y (50) values of fully after-ripened seeds for each seed the single variable y (50), holding h , s , and T
b
collection and incubation temperature.
b HT yb b
constant for the population. This approach does not give
the best fit for each individual curve, but creates a useful
are included in the retrieval plots for Whiterocks, general model ( Table 2; Fig. 1).
Hobblecreek, and Potosi Pass. The Strawberry seed collec- The modelling process is based on the decision that
tion ripened later than the other collections because of changes in y (50) through time at different storage tem-
b
the subalpine habitat ( Table 1). For this reason, only 4 peratures can be represented using thermal after-ripening
1242 Bauer et al.
time. This requires the assumption that changes in y (50) To incorporate effects of higher soil water potentials,
b
through time at any storage temperature are approxi- an upper threshold value for after-ripening was chosen.
mately linear. This assumption contradicts the observa- No changes in hydrothermal time parameters due to
tion in Christensen et al. (1996) that changes in y (50) water potentials above the threshold (i.e. from rainfall )
b
during after-ripening at 20 °C were not linear, but were are included in the model. During early stages of after-
more rapid early during after-ripening. However, with ripening, seeds progress relatively little toward germina-
the considerably greater data set available for the present tion during periods of high soil water potential, because
study (133 data points versus eight in Christensen et al., of their high y (50) values. During later stages of after-
b
1996), it was concluded that there is no basis for con- ripening, seeds accumulate greater progress toward ger-
sidering this relationship to be non-linear. mination during periods of high soil water potential,
Using a repeated regression analysis that incorporated confounding the after-ripening effect ( Fig. 3). Once
data points from later in the after-ripening process gave obvious progress toward germination has occurred (e.g.
a better fit to the field observations than a thermal time spurious values of y (50) for field retrievals, observed
b
model based on interpolation of halfway times (data not germination in the field ), the after-ripening model can no
shown). In a previous paper interpolated mean time longer be applied. Extension of this model to predict field
indices were used to characterize changes in three other germination would require application of a method to
indicators of dormancy loss in B. tectorum (dormant seed account for accumulation of hydrothermal time by par-
percentage, mean germination time of the germinable tially after-ripened as well as fully after-ripened seeds.
fraction, and synchrony index; Allen et al., 1995). Mean Model building resulted in several insights into charac-
or halfway times were then plotted for these three indices teristics of B. tectorum seed populations. The y (50)
b
against temperature and fitted the relationships with values for recently harvested seeds generally varied among
negative exponential equations. Inverse functions and populations at each incubation temperature ( Table 2).
negative exponential functions generate very similar The initial y (50) value largely determined how quickly
b
curves. It is therefore likely that the relationships between seeds reached fully after-ripened status in the field.
rate of change during storage in those dormancy indices Comparison of the field after-ripening plots ( Fig. 3)
could also have been fit successfully with a thermal after- shows that after-ripening of seeds in the field occurred at
ripening time model. Similarly, the mean dormancy period a relatively constant rate across seed collections and
in six cultivars of rice was shown to be a negative log- incubation temperatures. This uniformity in slopes is
linear function of storage temperature (see Roberts, 1965; largely due to the fact that seed collections after-ripened
and Fig. 7 in Roberts, 1988), suggesting that a thermal at similar rates at the high temperatures prevailing during
after-ripening time model could be applied to those the summer of 1995. Lower field seed zone temperatures
data as well. The dormancy loss model presented in would have resulted in more divergent slopes due to an
Christensen et al. (1996) and extended here incorporates accentuation of the effect of differences in base after-
the effects observed in earlier work (Allen et al., 1995), ripening temperatures.
as the three dormancy indices used in that work change Constancy or variation of y (50) values across incuba-
b
simultaneously in concert with underlying changes in tion temperatures gives an indication of the optimum
y (50). germination temperature. When two incubation temper-
b
Model development also required decisions on how atures are both below optimum, differences in germination
seed zone temperature and water potential could best be can be accounted for by thermal time only, and y (50)
b
used to predict field after-ripening. The model contains values are approximately the same for both incubation
the assumption that the effect of fluctuating seed zone temperatures. When at least one incubation temperature
temperatures is equivalent to a summation of the effects is above optimum, differences in germination between
of short constant temperature intervals. Another decision incubation temperatures cannot be explained by thermal
was to ignore the effect of water potential over the dry time only, and require variation in y (50) values for
b
after-ripening range. Previous studies have shown that explanation. The constancy of y (50) values across the
b
seed water potential can affect after-ripening rates incubation temperatures 15 °C and 25 °C ( Table 2), sug-
(Leopold et al., 1988; Foley, 1994). The laboratory data, gests that fully after-ripened seeds of the Whiterocks,
based on storage over a range of saturated salt solutions, Hobblecreek and Strawberry collections have an optimum
show that B. tectorum seeds after-ripen at a uniform rate temperature above 25 °C. Variation in y (50) values
b
over a wide range of water potentials above −150 MPa, between incubation temperatures shows that recently har-
and only exhibit delayed after-ripening at water potentials vested seeds of all collections, and fully after-ripened
below this threshold (unpublished data). The success of seeds of the Potosi Pass collection, have an optimum
this temperature-only model indicates that seed water temperature below 25 °C.
potentials were generally above −150 MPa during dry The y (50) values of fully after-ripened seeds were
b
after-ripening in the field. apparently not affected by whether incubation temper-
Modelling seed dormancy loss 1243
atures with the same mean were constant or alternating, Bouwmeester HJ, Karssen CM. 1992. The dual role of
but this was not true for recently harvested seeds (Table 2). temperature in the regulation of the seasonal changes in
dormancy and germination of seeds of Polygonum persicaria
The difference in effect between recently harvested and L. Oecologia 90, 88–94.
fully after-ripened seeds may have resulted from decreased Bradford KJ. 1990. A water relations analysis of seed germina-
requirements for specific incubation conditions (e.g. tion rates. Plant Physiology 94, 840–9.
alternating temperatures) in after-ripened seeds. Bradford KJ. 1995. Water relations in seed germination. In:
A major limitation of this thermal after-ripening time Kigel J, Galili G, eds. Seed development and germination.
New York: Marcel Dekker, Inc., 351–96.
model in its present form is the fact that each incubation
Bradford KJ. 1996. Population-based models describing seed
temperature requires a unique set of parameters. It is dormancy behaviour: implications for experimental design
tempting to speculate that a more general thermal after- and interpretation. In: Lang GA, ed. Plant dormancy:
ripening time model could be built by assuming that the physiology, biochemistry, and molecular biology. Wallingford:
base temperature for after-ripening and the slope of the CAB International, 313–39.
change in y (50) are constant across incubation temper- Christensen M, Meyer SE, Allen PS. 1996. A hydrothermal
b time model of seed after-ripening in Bromus tectorum L. Seed
atures. This would mean that the difference between the Science Research 6, 1–9.
initial and final y (50) values would be the only variable D’Antonio CM, Vitousek PM. 1992. Biological invasions by
b
affecting the value of h at each incubation temperature. exotic grasses, the grass/fire cycle, and global change. Annual
AT
It has been demonstrated that this more general model Review of Ecological Systems 23, 63–87.
applies to the after-ripening process in another grass Ellis RH, Simon G, Covell S. 1987. The influence of temperature
on seed germination rate in grain legumes. III. A comparison
species of the sagebrush steppe (Elymus elymoides, bottle-
of five faba bean genotypes at constant temperature using a
brush squirreltail; Debaene et al., unpublished results). new screening method. Journal of Experimental Botany
Unfortunately, analysis of covariance for B. tectorum 38, 1033–43.
collections in the present study, with incubation temper- Foley ME. 1994. Temperature and water status of seed affect
ature as the class variable and thermal time in storage as afterripening in wild oat (Avena fatua). Weed Science
the continuous variable, indicates that the slopes of change 42, 200–4.
Gummerson RJ. 1986. The effect of constant temperatures and
in y (50) at different incubation temperatures are not
b osmotic potentials on the germination of sugar beet. Journal
identical, whether or not the model is constrained to a of Experimental Botany 37, 729–41.
common base temperature for after-ripening (data not Hanks RJ. 1992. Applied soil physics. New York: Springer-
shown). The apparently complex relationship between Verlag.
incubation temperature and after-ripening rate in this Hull AC, Hansen WT. 1974. Delayed germination of cheatgrass
seed. Journal of Range Management 27, 366–8.
species will require clarification with data from a more
Karssen CM, Derkx MPM, Post BJ. 1988. Study of seasonal
comprehensive set of incubation temperatures. variation in dormancy of Spergula arvensis L. seeds in a
condensed annual temperature cycle. Weed Research 28,
449–57.
Acknowledgements Kirchner TB. 1990. Time-zero: the integrated modeling environ-
ment. Fort Collins, Colorado: Quaternary Software, Inc.
This research was funded in part from USDA CSREES grant
Leopold AC, Glenister R, Cohn MA. 1988. Relationship between
number 93–38300–8688. Assistance with technical work was
water content and afterripening in red rice. Physiologia
provided by Jennifer Baker, Dedra Cutler and Stephanie
Carlson. We acknowledge Kent Bradford for a substantively Plantarum 74, 659–62.
helpful review of an earlier draft of this manuscript. Meyer SE. 1992. Habitat-correlated variation in firecracker
penstemon (Penstemon eatonii: Scrophulariaceae) seed ger-
mination response. Bulletin of the Torrey Botanical Club
119, 268–79.
References Meyer SE, Allen PS. 1995. Ecological genetics of seed
Allen PS, Meyer SE. 1990. Temperature requirements for seed germination regulation in Bromus tectorum L. Abstract.
germination of three Penstemon species. HortScience 25, Supplement of the Bulletin of the Ecological Society of America
191–3. 76, 183.
Allen PS, Meyer SE. 1998. Ecological aspects of seed dormancy Meyer SE, Allen PS, Beckstead J. 1997. Seed germination
loss. Seed Science Research 8, (in press). regulation in Bromus tectorum L. (Poaceae) and its ecological
Allen PS, Meyer SE, Beckstead J. 1995. Patterns of seed after- significance. Oikos 78, 475–85.
ripening in Bromus tectorum L. Journal of Experimental Meyer SE, Kitchen SG. 1992. Cyclic seed dormancy in the
Botany 46, 1737–44. short-lived perennial Penstemon palmeri. Journal of Ecology
Baskin JM, Baskin CC. 1985. The annual dormancy cycle in 80, 115–22.
buried weed seeds: a continuum. BioScience 35, 492–8. Meyer SE, Kitchen SG. 1994a. Habitat-correlated variation in
Beckstead J, Meyer SE, Allen PS. 1996. Bromus tectorum seed seed germination response to chilling in Penstemon Section
germination: between-population and between-year variation. Glabri (Scrophulariaceae). American Midland Naturalist
Canadian Journal of Botany 74, 875–82. 132, 349–65.
Benech-Arnold RL, Sanchez RA. 1995. Modeling weed seed Meyer SE, Kitchen SG. 1994b. Life history variation in blue
germination. In: Kigel J, Galili G, eds. Seed development and flax (Linum perenne: Linaceae): seed germination phenology.
germination. New York: Marcel Dekker, 545–66. American Journal of Botany 81, 528–35.
1244 Bauer et al.
Meyer SE, McArthur ED, Jorgensen GL. 1989. Variation in Roberts EH. 1965. Dormancy in rice seed. IV. Varietal responses
germination response to temperature in rubber rabbitbrush to storage and germination temperature. Journal of
(Chrysothamnus nauseosus: Asteraceae) and its ecological Experimental Botany 16, 341–9.
implications. American Journal of Botany 76, 981–91. Roberts EH. 1988. Temperature and seed germination. In: Long
Meyer SE, Monsen SB. 1992. Big sagebrush germination SP, Woodward FI, eds. Plants and temperature. Cambridge:
patterns: subspecies and population differences. Journal of Company of Biologists, 109–32.
Range Management 45, 87–93. Wicks GA, Burnside OC, Fenster CR. 1971. Influence of soil
Meyer SE, Monsen SB, McArthur ED. 1990. Germination type and depth of planting on downy brome seed, Weed
response of Artemisia tridentata (Asteraceae) to light and Science 19, 82–6.
chill: patterns of between-population variation. Botanical Wilson AM. 1973. Responses of crested wheatgrass seed to
Gazette 151, 176–83. environment. Journal of Range Management 26, 43–6.
Milby TH, Johnson FL. 1987. Germination of downy brome Young JA, Evans RA. 1975. Germinability of seed reserves in a
from southern Kansas, central Oklahoma, and north Texas. big sagebrush (Artemisia tridentata) community. Weed Science
Journal of Range Management 40, 534–6. 23, 358–64.