American Journal of Plant Sciences, 2011, 2, 134-147
doi:10.4236/ajps.2011.22015 Published Online June 2011 (http://www.SciRP.org/journal/ajps)
Switchgrass (Panicum virgatum L.) Intraspecific
Variation and Thermotolerance Classification
Using in Vitro Seed Germination Assay
Ramdeo Seepaul1, Bisoondat Macoon2, K. Raja Reddy1*, Brian Baldwin1
1Department of Plant and Soil Sciences, Mississippi State University, Mississippi, USA; 2Central Mississippi Research and Extension
Center, Raymond, Mississippi, USA.
*
Received March 4th, 2011; revised May 9th, 2011; accepted May 17th, 2011.
ABSTRACT
Cardinal temperatures for plant processes have been used for thermotolerance screening of genotypes, geoclimatic
adaptability determinatio n and phenological prediction. Curren t simulation models for switchgrass (Panicum virgatum
L.) utilize single cardinal temperatures across genotypes for both vegetative and reproductive processes although in-
tra-specific variation exists among genotypes. An experiment was conducted to estimate the cardinal temperatures for
seed germination of 14 diverse switchgrass genotypes and to classify genotypes for temperature tolerance. Stratified
seeds of each genotype were germinated at eight constant temperatures from 10˚C to 45˚C under a constant light inten-
sity of 35 µmo l·m–2·s–1 for 12 h·d–1. Germination was recorded at 6-h intervals in all treatmen ts. Maximum seed germi-
nation (MSG) and germination rate (GR), estimated by fitting Sigmoidal function to germina tion-time series data, var-
ied among genotypes. Quadratic and bilinear models best described the MSG and GR responses to temperature, re-
spectively. The mean cardinal temperatures, Tmin, Topt, and Tmax, were 8.1, 26.6, and 45.1˚C for MSG and 11.1, 33.1,
and 46.0˚C for GR, respectively. Cardinal temperatures for MSG and GR; however, varied significantly among geno-
types. Genotypes were classified as sensitive (‘Cave-in-rock’, ‘Dacotah’, ‘Expresso’, ‘Forestburg’, ‘Kanlow’, ‘Sun-
burst’, ‘Trailblazer’ and ‘Tusca’), intermediate (‘Alamo’, ‘Blackwell’, ‘Carthage’, ‘Shawnee’, and ‘Shelter’) and tol-
erant (‘Summer’) to high temperature based on cumulative temperature response index (CTRI) estimated by summing
individual response indices estimated from the MSG and GR cardinal temperatures. Similarly, genotypes were also
classified as sensitive (Alamo, Blackwell, Carthage, Dacotah, Shawnee, Shelter and Summer), moderately sensitive
(Cave-in-rock, Forestburg, Kanlow, Sunburst, and Tusca), moderately tolerant (Trailblazer), and tolerant (Expresso) to
low temperatures. The cardinal temperature estimates would be useful to improve switchgrass models for field applica-
tions. Additionally, the identified cold- and heat-tolerant genotypes can be selected for niche environments and in
switchgrass breeding programs to develop new genotypes for low and high temperature environments.
Keywords: Switchgrass, Cardinal Temperature, Temperature Tolerance, Germination, Genotype Variability, Response
Index, Screening, Genotype Classification
1. Introduction
The adoption of a biomass feedstock crop for a niche
environment is favoured on the species ability to grow
and sustain under a wide range of growing conditions
and its ability to produce high yields and quality biomass.
From an agronomic perspective, the crop should also be
able to establish rapidly and uniformly under existing
conditions to escape weed competition and late-season
water unavailability [1]. Establishment of warm-season
feedstock grasses has been limited due to slow germina-
tion and low seedling vigor [2,3], particularly in the first
year after seeding, presenting a major problem in the
improvement of existing stands, or in establishing new
stands. Slight or moderate successes of native grasses
establishment can be attributed to seed dormancy and
delayed germination [4]. Seeding feedstock fields re-
quires knowledge of many parameters, including opti-
mum temperature and moisture conditions for rapid ger-
mination and establishment [5,6].
Switchgrass (Panicum virgatum L.), a warm-season,
native C4 bunch grass species was identified as a poten-

Switchgrass (Panicum virgatum L.) Intraspecific Variation and Thermotolerance Classification 135
Using in Vitro Seed Germination Assay
tial and model lignocellulosic biofuel feedstock by the
U.S. Department of Energy’s Bioenergy Feedstock De-
velopment Program [7]. It is a highly diverse species with
significant genetic [8] and phenotypic variation resulting
from gene migration, random genetic drift, mutation,
natural selection [9] combined with environment dis-
similarity due to latitude, altitude, soil type, and precipi-
tation [10].
Temperature is a major environmental factor influ-
encing seed germination capacity and rate and seedling
vigor [3] through three distinct processes; it effects on
seed deterioration (seed aging), dormancy loss, and on
the germination process itself [11]. Extreme temperatures
are the single most important factor delimiting the dis-
tribution, adaptability, and yield potential of plants. Sub-
and supra-optimal soil temperatures at seeding can affect
both the germination rate and maximum seed germina-
tion; therefore breeding for seed temperature tolerance may
be necessary for adequate and uniform crop establishment.
Determining temperature effects on seed germination
using mathematical functions may be useful in evaluat-
ing germination characteristics or establishment potential
among genotypes or species [12]. Final seed germination
percentage and germination rate are both considered sen-
sitive indicators of seed vigor [13]. Germination can be
characterized by three cardinal temperatures (minimum,
Tmin; maximum, Tmax and optimum, Topt) that determine
the range of temperatures across which germination can
occur. Previous studies that reported effects of tempera-
ture on switchgrass germination capacity and rate did not
quantify the cardinal temperatures for diverse switch-
grass genotypes. Switchgrass germinates slowly when
the temperature is below 15.5˚C with maximum germi-
nation occurring within 3 d of imbibition at 29.5˚C [14].
Minimum temperature for switchgrass germination is
10.3˚C and optimum temperature occurring between 25˚C
and 30˚C [15]. Minimum temperatures are critical for
accurate phenological predictions because minute dif-
ferences in temperatures can cause considerable differ-
ences in germination time. Current switchgrass models
that simulate switchgrass phenology use blanket mini-
mum temperatures that range from 10˚C to 12˚C [16-18],
although it is suspected that there is intra-species variation.
The interest in switchgrass as a feedstock has fostered
development and selection of a wide number of geno-
types, which must be screened for various abiotic stress
tolerances prior to release. Current screening methods
are restricted to field performance and visual evaluations
which may mask a genotype’s true potential or tolerance
capacity due to unpredictable moisture and fluctuating
temperatures in the field. Field screening for temperature
tolerance is tedious, inconsistent, and seasonally limited;
therefore the need for simple, rapid, and reliable tech-
niques to identify sources of tolerance and for evaluating
a large number of breeding materials in controlled envi-
ronments is required [19]. Screening for abiotic stress
tolerance has been achieved using biochemical and
physiological parameters at the germination, emergence,
vegetative, and reproductive stages. In vitro seed-based
screening can provide insights into genotypic environ-
mental adaptability and tolerance capacity prior to field
evaluations. Studies related to temperature tolerance
screening in switchgrass; however, are limited in general
and no reported studies using seed-based parameters
have been found. Seed-based parameters, in particular,
germination capacity and rate have been used success-
fully to screen several species and genotypes for various
abiotic stress factors including drought [20,21], saline
[22,23], flooding/water logging [24], chilling [25,26],
and heat tolerance [27,28] in other species. The tem-
perature tolerance capacity of different genotypes may be
determined by relative ranking using single value indices,
percentiles and quartiles relative to control studies and
cumulative indices, groupings based on statistical sepa-
ration of means [28-30] or quantitative relationships de-
termined by principal component analysis [31-33].
The objectives of this study were to 1) quantify the
effects of temperature on seed germination capacity and
rate, 2) determine the cardinal temperatures for seed
germination capacity and rate, and 3) classify genotypes
for temperature tolerance using cumulative temperature
response index concept. The seed germination and tem-
perature dependent functional algorithms developed from
these data are a prerequisite for modeling the germina-
tion of switchgrass genotypes adapted to different cli-
matic zones.
2. Materials and Methods
2.1. Seed Material
Seeds of 14 switchgrass genotypes, representative of
northern and southern, upland and lowland ecotypes,
were evaluated in this experiment (Table 1). For nine
cultivars, seeds were collected from the plants grown
during the 2006-2007 growing season at Mississippi
State, MS (33°28′N, 88°47′W) and stored at 10˚C and
40% RH. Seeds of Blackwell, Carthage, Cave-in-Rock,
Shawnee, and Shelter were obtained from the Ernst Seed
Company (Meadville, PA) from the 2006-2007 growing
season and stored at similar conditions. All seeds were
kept in cold storage to maintain seed quality prior to
testing. Seeds were homogenously mixed and 100 seed
er experimental unit for germination testing were counted p

Switchgrass (Panicum virgatum L.) Intraspecific Variation and Thermotolerance Classification
Using in Vitro Seed Germination Assay
136
Table 1. Ploidy level, ecotype, latitude, origin and plant hardiness zone (PHZ) of switchgrass ge notypes.
Genotype Ploidy
Level EcotypeLatitude Origin PHZ Remarks Reference
Alamo T lowland southern TX 6 Selected for biomass
Blackwell H upland S Blackwell, OK 5a Riley and Vogel (1982)
Carthage O upland southern IL
Cave-in-Rock H lowland/
upland S Cave-in-Rock, IL 4b Riley and Vogel (1982)
Dacotah T upland North Dakota 4a
Early maturity, winter
hardy, high stand density,
persistent
Barker et al. (1990)
Expresso lowland Mississippi Selected for improved
germination
Forestburg T upland N Forestburg, SD 3b-4b
Early, maturity, excellent
winter hardiness and
persistence, good seed
potential
Barker et al. (1988)
Kanlow T lowlandN Wetumka, OK 5
Shawnee O upland S Cave-in-Rock, IL High forage yield and
quality Vogel et al. (1996)
Shelter H
lowland/
upland N St. Mary’s, WV 4 Wullschleger et al. (1996)
Summer T upland Southern NE 4
Sunburst H upland N South Dakota
Winter hardy, leafy,
heavy-seeded, superior
seedling vigor
Boe and Ross (1998);
Wullschleger et al. (1996)
Trailblazer H upland N Nebraska High forage quality, high
IVDMD Vogel et al. (1991)
Tusca lowland Mississippi
Selected for herbicide
tolerance from Alamo
Genotypes are classified based on ploidy level (T = tetraploid, H = hexaploid, and O = octaploid), and latitude of adaptation (S = southern and N = Northern).
by an electronic seed counter (Model 850-2; The Old
Mill Company, Savage, MD).
2.2. Seed Germination Testing
Stratified seeds (14 d at 5˚C) were used for germination
testing from March to May 2009 according to Associa-
tion of Official Seed Analysts (ASOA) rules with no
humidity control. Seeds were blotted and placed imme-
diately to the testing temperature to minimize drying
which induces secondary dormancy [34].
Preliminary studies at low temperature (<20˚C) indi-
cated that fungal infection can affect germination,
prompting the use of Captan{cis-N-[(trichloromethyl)th-
io-4-cyclohexene-1,2- dicarboximide]} at 0.55 g·ai·kg–1
seed as a drench prior to germination testing at all tem-
peratures. Each genotype was replicated four times in a
completely randomized design with 100 seed per repli-
cate placed on a moistened single layer Whatman No. 1
filter paper (Whatman, Atlanta, GA) in a covered 90-cm
sterilized disposable plastic Petri dish to minimize mois-
ture loss. Petri dishes were vertically stacked at constant
set temperatures, 10 to 45˚C at 5˚C intervals. Constant
light with a photon flux density of 35 ± 2.6 µmol·m2·s–1
was provided by cool white fluorescent lamps during a
12-h light period, for all genotypes and temperatures in
five germination chambers (Fisher Scientific, Suwanee,
GA). Petri dishes were monitored daily to ensure that the
filter paper remained moist and watered when necessary
with distilled water.
Replicates for each genotype were completely ran-
domized within the germination chamber for each tem-
perature. To minimize the potential of small temperature
changes within the chambers, the Petri dishes were rear-
ranged every 6 h. Germinated seeds were counted, re-
corded and discarded every 6 h. Counts were discontin-
ued if no seed germinated for five consecutive days. A
seed was considered germinated when the coleoptile or
coleorhizae was at least 2 mm long.

Switchgrass (Panicum virgatum L.) Intraspecific Variation and Thermotolerance Classification 137
Using in Vitro Seed Germination Assay
2.3. Curve Fitting Procedure and Data Analysis
Temperature and germination time-course data were fit-
ted with a 3-parameter Sigmoidal function (Equation (1))
using SigmaPlot 11 [35]. This function estimated the
maximum cumulative seed germination percentage (ger-
mination capacity); the shape and steepness of the curve;
and time to reach 50% of maximum germination. The
rate of development was derived by the reciprocal of
time to 50% of maximum seed germination.


max50 rate
1expGGxx G



(1)
where G is the total seed germination percentage, Gmax is
the maximum cumulative seed germination percentage,
x50 is the time to 50% maximum seed germination, and
Grate is the slope of the curve.
Maximum seed germination and germination rate re-
sponses to temperature were analyzed using linear and
nonlinear regression techniques for all genotypes [31].
Based on the highest coefficient of determination (r2)
value and the root mean square error (RMSE), the best
curve fitting model was obtained. Accordingly, maxi-
mum seed germination was modeled using a quadratic
function (r2 = 0.88, RMSE = 5.2) while germination rate
was modeled by a modified bilinear function (r2 = 0.95,
RMSE = 1.00). Quadratic and modified bilinear equa-
tions estimates for each replicate within each genotype
were estimated using PROC NLIN of SAS [36] with a
modified Newton Gauss iterative method. For the quad-
ratic model (Equation (2)), the three cardinal tempera-
tures (Tmin, Topt and Tmax), were estimated using Equation
(3) to (5).
2
MSGT Tab c  (2)
opt
T(2)bc (3)

2
Tbb 
min42acc
(4)

2
max
T4bbac 2c

(5)
where MSG is the maximum seed germination, Topt, Tmin
and Tmax are the optimum, minimum, and maximum car-
dinal temperatures for seed germination, respectively, T
is treatment temperature at which MSG was determined,
and a, b, and c are genotype-specific constants generated
using PROC GLM in SAS [36]. For the modified bilin-
ear model using Equation (6), Topt was generated using
SAS [36] while Tmin and Tmax were estimated using
Equation (7) and (8).
 
1opt 2opt
GRTTABS TTab b (6)


min2 1opt12
TTabb bb


 

(7)


max2 1opt1 2
TTabb bb


 

(8)
where GR is germination rate, Topt, Tmin, and T
max is the
optimum, minimum, and maximum cardinal tempera-
tures for seed germination, respectively, T is the treat-
ment temperature, and a, b1 and b2 are genotype-specific
constants generated using PROC NLIN in SAS [36].
2.4. Cumulative Temperature Response Index
(CTRI)
Switchgrass genotypes were classified into cold or heat
tolerant groups based on the summation of individual
temperature response index values following the protocol
used by Salem et al. [30] for pollen germination response
to temperature. Accordingly, heat CTRI (H-CTRI) was
calculated as the MSG and GR values for each of the
cardinal temperatures (Tmin, Topt and Tmax) of a specific
genotype, divided by the maximum value observed
among all genotypes (Equation [9]) while cold CTRI
(C-CTRI) was determined by dividing the minimum
value among all genotypes by the value of a specific
genotype (Equation [10]), where h and t refers to maxi-
mum and genotype-specific parameter values. Genotypes
were classified based on CTRI of all parameters as cold-
tolerant (>minimum CTRI + 4 standard deviations [SD]),
moderately cold-tolerant (>minimum CTRI + 3 SD),
moderately cold-sensitive (>minimum CTRI + 2 SD),
and cold-sensitive (>minimum CTRI + 1 SD). Similarly,
genotypes were classified as heat-sensitive (>minimum
CTRI + 1 SD), intermediate (>minimum CTRI + 2 SD),
and heat tolerant (>minimum CTRI + 3 SD).
All cumulative germination data were arcsine trans-
formed prior to analysis and back transformed for re-
porting. Replicated values of cardinal temperatures (Tmin,
Topt, and Tmax), temperature adaptability range (TAR =
Tmax – Tmin), and MSG were analyzed using the ANOVA
procedure (PROC GLM) in SAS [36] to determine the
tt
hh
tt t
hh h
min optmax
min optmax
min opt max
min opt max
MSG TMSG TMSG T
H-CTRI MSG TMSG TMSG T
GR TGR TGR T
GR TGR TGR T

 






t
h
(9)
hh
tt
hh h
tt t
min opt max
min optmax
min opt max
min opt max
MSG TMSG TMSG T
C-CTRI MSG TMSG TMSG T
GR TGR TGR T
GR TGR TGR T









h
t
(10)

Switchgrass (Panicum virgatum L.) Intraspecific Variation and Thermotolerance Classification
Using in Vitro Seed Germination Assay
138
effect of temperature treatment on MSG and GR and
their respective cardinal temperatures (Tmin, Topt, and
Tmax). Cardinal temperatures for MSG and GR parameter
means were separated using Fishers protected least sig-
nificant differences (LSD) at P = 0.05. Germination pa-
rameters (MSG and GR) were treated as dependent vari-
ables while temperature and time to germination as in-
dependent variables. Regression analysis was carried out
using SigmaPlot 11.0. Also, the mean seed germination
parameters response to temperature was tested based on
lowland (Alamo, Expresso, Kanlow and Tusca) or up-
land (Blackwell, Carthage, Cave-in-Rock, Dacotah, For-
estburg, Shawnee, Shelter, Summer, Sunburst and Trail-
blazer) ecotypes using Fishers protected least significant
differences (LSD) at P = 0.05.
variability of genotypes in their germination characteris-
tics (Figure 1). For clarity, only data and fitted lines for
four genotypes, each representative of northern and
southern upland (Cave-in-Rock and Shelter) and lowland
(Alamo and Kanlow) ecotypes are presented. There was
no germination at 10 or at 45°C in any of the genotypes
tested.
3.2. Maximum Seed Germination Response to
Temperature
Among the linear and nonlinear regression models tested,
the quadratic function best described the response of
MSG to temperature (mean r2 = 0.93, RMSE = 5.2). For
clarity, only data and fitted lines for four genotypes, each
representative of Northern and southern upland (Cave-
in-Rock and Shelter) and lowland (Alamo and Kanlow)
genotypes are presented (Figure 2). Maximum seed
germination varied (P < 0.001) among genotypes with a
mean of 73% and ranged from 41 (Alamo) to 93% (Ex-
presso) (Table 2). Cardinal temperatures (Tmin, Topt, and
Tmax) for MSG also differed among the genotypes (P
3. Results
3.1. Germination Time Courses
The 3-parameter Sigmoidal function fitted the cumula-
tive germination time course (mean r2 = 0.98) of geno-
types response to temperature efficiently, illustrating the
Figure 1. Germination time courses for seeds of (a) Alamo, (b) Cave-in-Rock, (c) Kanlow and (d) Shelter switchgrass
germinated at a range of temperature (15˚C - 40˚C). The symbols indicate the observed cumulative germination data and the
lines indicate the germination time courses fitted using a three-parameter sigmoidal function. Data are means and ± SE of
four replications.

Switchgrass (Panicum virgatum L.) Intraspecific Variation and Thermotolerance Classification 139
Using in Vitro Seed Germination Assay
Figure 2. Influence of temperature on maximum seed germination and along with the fitted quadratic equations of four
switchgrass genotypes (Alamo, Kanlow, Shelter, and Cave-in-Rock). The symbols are recorded maximum germination
percentages and the curves are fitted lines using quadratic functions.
Table 2. Maximum seed germination percentage (MSG), temperature adaptability range (TAR), quadratic equation con-
stants (a, b, and c), regression coefficients (r2), and cardinal temperatures (Tmin, Topt, and Tmax) for maximum seed germina-
tion (MSG) of 14 switchgrass genotypes in response to temperature.
Equation constants r2 Cardinal temperatures (oC)
Genotype MSG (%) TAR (°C) a b c Tmin T
opt T max
Alamo 40.97 ± 1.56 34.94 ± 0.14 –46.48 5.88 –0.10850.859.61 ± 0.19 27.08 ± 0.12 44.55 ± 0.05
Blackwell 83.23 ± 2.16 36.01 ± 0.25 –119.0315.28–0.27980.989.33 ± 0.32 27.34 ± 0.20 45.34 ± 0.10
Carthage 55.09 ± 1.39 35.51 ± 0.44 –80.43 9.68 –0.17330.9310.2 ± 0.35 27.95 ± 0.13 45.71 ± 0.11
Cave-in-Rock 79.48 ± 1.38 40.53 ± 1.03 –31.75 9.02 –0.17990.905.62 ± 0.96 25.88 ± 0.45 46.14 ± 0.10
Dacotah 85.68 ± 3.36 34.25 ± 0.39 –124.8715.25–0.27860.9710.4 ± 0.41 27.52 ± 0.22 44.64 ± 0.09
Expresso 93.07 ± 0.55 43.38 ± 0.62 –41.99 11.12–0.21760.793.69 ± 0.48 25.38 ± 0.18 47.07 ± 0.16
Forestburg 80.76 ± 2.72 37.26 ± 0.23 –72.49 11.38–0.21720.957.68 ± 0.13 26.31 ± 0.10 44.95 ± 0.17
Kanlow 53.05 ± 6.74 37.95 ± 1.09 –29.15 5.52 –0.10980.926.40 ± 0.97 25.37 ± 0.43 44.34 ± 0.15
Shawnee 50.31 ± 1.85 35.41 ± 0.26 –74.79 9.25 –0.16750.989.90 ± 0.26 27.60 ± 0.14 45.31 ± 0.05
Shelter 74.27 ± 2.39 33.47 ± 0.20 –118.7213.04–0.23130.9411.46 ± 0.21 28.19 ± 0.12 44.92 ± 0.08
Summer 67.52 ± 1.32 31.47 ± 0.27 –151.2014.61–0.25250.9512.83 ± 0.11 28.56 ± 0.09 44.30 ± 0.21
Sunburst 86.95 ± 0.21 40.65 ± 1.75 –60.75 11.39–0.22130.985.49 ± 1.07 25.81 ± 0.38 46.14 ± 0.82
Trailblazer 87.46 ± 1.98 41.78 ± 0.94 –42.23 10.63–0.21140.944.19 ± 0.84 25.08 ± 0.37 45.97 ± 0.14
Tusca 89.56 ± 0.78 35.54 ± 1.33 –76.87 12.88–0.24300.906.27 ± 0.82 24.04 ± 0.48 41.81 ± 0.82
Mean 73.39 37.01 − − − 0.938.08 26.58 45.09
LSD 12.66* 4.09* − − − 3.09* 1.43* 1.70*
*Significant at P = 0.05 probability level.

Switchgrass (Panicum virgatum L.) Intraspecific Variation and Thermotolerance Classification
Using in Vitro Seed Germination Assay
140
3.3. Germination Rate Response to Temperature
< 0.001). The Tmin values ranged from 3.69 (Expresso) to
12.83˚C (Summer) with a mean of 8.08˚C. The Topt was
26.58˚C (Table 2); however, there was variation among
the genotypes (P < 0.001). Summer recorded the highest
Topt (28.56˚C) while Tusca showed the lowest (24.04˚C).
The Tmax ranged from 41.81 (Tusca) to 47.07˚C (Ex-
presso) with a mean of 45.07˚C (Table 2). The TAR for
MSG ranged from 43.38 (Expresso) to 31.37˚C (Summer)
with a mean of 37˚C for all genotypes.
The modified bilinear equation best described the rela-
tionship between GR and temperature (mean r2 = 0.95,
RMSE = 1.0) among the linear and non-linear models
tested. Cardinal temperatures for GR differed among
genotypes (P < 0.05) (Table 3). For clarity, only data
and predictor lines of four genotypes are presented in
Figure 3. The Tmin ranged from 9.09 (Dacotah) to
12.92˚C (Shelter) with a mean of 11.13˚C. A mean of
33.12˚C was estimated for Topt which ranged from 29.55
(Shelter) to 35.73˚C (Tusca). Highest Tmax was recorded
in Shelter (48.15˚C), while the lowest Tmax (45.0˚C) was
observed in Kanlow. The TAR ranged from 32.92
(Blackwell) to 36.18˚C (Dacotah) with a mean of
34.88˚C (Table 3). Ecotypic classification of genotypes
indicate that TAR, Tmin and Tmax did not differ, but Topt
was different (P < 0.05) with a mean of 32.37 and
34.98˚C for upland and lowland ecotypes, respectively
(P = 0.0477; LSD = 2.57). Cardinal temperatures variation
was small between ecotypes (<4%) with germination rate
Tmin being more variable than Topt and Tmax for both
upland and lowland ecotypes (Table 3).
Grouping genotypes based on upland and lowland
ecotype revealed no differences (P > 0.05) for MSG,
TAR, Tmin and Tmax; however, Topt for MSG was differ-
ent (P = 0.0471, LSD = 1.53) with mean of 27.02 and
25.47˚C for upland and lowland ecotypes, respectively.
Maximum seed germination for both upland and lowland
ecotypes also varied ( 10%) (data not shown). Cardinal
temperature (Tmin, Topt and Tmax) variation was small
between ecotypes (<4%). Maximum seed germination
Tmin was more variable than Topt and Tmax for both upland
and lowland ecotypes. On average, MSG cardinal
temperatures were 10 and 6% more variable than germina-
tion rate cardinal temperatures for upland and lowland
ecotypes, respectively.
Figure 3. Effect of temperature on germination rate along with the fitted modified bilinear fitted lines and equations of four
switchgrass genotypes (Alamo, Kanlow, Shelter, and Cave-in-Rock). The symbols are the derived germination rate and the
lines are predicted values by the fitted modified bilinear equations.

Switchgrass (Panicum virgatum L.) Intraspecific Variation and Thermotolerance Classification 141
Using in Vitro Seed Germination Assay
Table 3. Temperature adaptability range (TAR), modified bilinear equation constants (a, b, and c), regression coefficients (r2),
and cardinal temperatures (Tmin, Topt, and Tma x) for germination rate of 14 switchgrass genotypes in response to temperature.
Equation Constants Cardinal temperatures (oC)
GeNotype TAR (˚C)
a b c
r2
Tmin T
opt T max
Alamo 34.29 ± 0.86 0.5255 –0.0094 –0.0334 0.95 11.96 ± 0.60 33.02 ± 1.40 46.25 ± 0.78
Blackwell 32.92 ± 0.26 0.6791 –0.0142 –0.0459 1.00 12.14 ± 0.21 33.91 ± 0.08 45.06 ± 0.07
Carthage 34.06 ± 0.47 0.5945 0.0010 –0.0349 0.87 12.83 ± 0.22 30.45 ± 1.00 46.89 ± 0.64
Cave-in-Rock 35.11 ± 0.57 0.6430 –0.0282 –0.0509 0.98 10.16 ± 0.60 34.43 ± 0.88 45.27 ± 0.37
Dacotah 36.18 ± 0.30 0.6469 –0.0266 –0.0496 0.97 9.09 ± 0.49 35.34 ± 0.88 45.27 ± 0.26
Expresso 35.77 ± 0.53 0.7545 –0.0290 –0.0566 0.98 9.33 ± 0.63 35.50 ± 0.83 45.09 ± 0.10
Forestburg 35.17 ± 0.44 0.5884 –0.0121 –0.0374 0.98 10.18 ± 0.52 34.03 ± 0.78 45.35 ± 0.12
Kanlow 35.06 ± 0.84 0.6227 –0.0196 –0.0453 1.00 9.94 ± 0.84 35.65 ± 0.26 45.00 ± 0.00
Shawnee 35.01 ± 0.45 0.5940 0.0024 –0.0338 0.82 12.54 ± 0.26 30.56 ± 1.05 47.55 ± 0.71
Shelter 35.23 ± 0.15 0.5661 0.0023 –0.0326 0.87 12.92 ± 0.08 29.55 ± 0.07 48.15 ± 0.10
Summer 35.02 ± 0.36 0.4765 0.0009 –0.0270 0.86 12.06 ± 0.49 30.77 ± 1.36 47.08 ± 0.68
Sunburst 35.35 ± 0.34 0.6072 –0.0008 –0.0343 0.89 11.21 ± 0.23 30.48 ± 1.04 46.55 ± 0.50
Trailblazer 35.59 ± 0.18 0.7006 –0.0273 –0.0524 0.97 9.86 ± 0.27 34.21 ± 0.39 45.44 ± 0.15
Tusca 33.52 ± 0.39 0.6361 –0.0089 –0.0384 0.90 11.65 ± 0.34 35.73 ± 1.02 45.16 ± 0.22
Mean 34.88 - - - 0.93 11.13 33.12 46.01
LSD 2.47* - - - - 2.32* 4.49* 2.17*
*Significant at P = 0.05 probability level.
3.4. Genotype Classification Using Cumulative
Temperature Response Index (CTRI)
Six parameters (Tmin, Topt, and Tmax for both MSG and
GR) were used for both heat- and cold-tolerance classi-
fication of genotypes based on CTRI. Each parameter
contributed differently based on its relation to the mini-
mum or maximum value for that parameter across the
genotypes. Using one standard deviation permitted the
classification of heat-CTRI values (which ranged from
4.83 to 6.05) into three groups (heat-sensitive [4.83 -
5.43]; intermediate [5.44 - 5.74], and heat-tolerant [5.73 -
6.05]). Summer was identified as the most heat-tolerant
genotype while Cave-in-Rock, Dacotah, Expresso, For-
estburg, Kanlow, Sunburst, Trailblazer and Tusca as
heat-sensitive genotypes (Tab le 4).
Using the same parameters used for heat tolerance, the
genotypes were similarly classified for cold-tolerance
(Table 4). Cold-CTRI values, which ranged from 4.74 to
6.21, allowed grouping of genotypes into four tolerance
categories (cold sensitive [4.74 - 5.03]; moderately
cold sensitive [5.04 - 5.32], moderately cold tolerant
[5.33 - 5.62], and cold tolerant [5.63 - 6.21]). Expresso
had the highest cold-CTRI (5.64), and therefore consid-
ered as most cold-tolerant genotype, while Summer had
the lowest cold-CTRI (4.74) and was classified as
cold-sensitive genotype (Table 4).
3.5. Parameter Relationships
No correlation was found between MSG Tmin and Tmax
and Topt and Tma x (P > 0.05), however, a positive linear
correlation existed between Tmin and Topt (r2 = 0.81, P <
0.0001). As GR Tmin increased among the genotypes,
Tmax generally increased (r2 = 0.56, P < 0.0021). An
inverse relationship was found between GR Tmin and Topt
(r2 = 0.58, P < 0.0014) as well as Topt and Tmax (r2 = 0.88,
P < 0.0001). The correlation between MSG and GR car-
dinal temperatures varied, but a weak positive correlation
was found between MSG and GR Tmin (r2 = 0.39, P =
0.0163), while a weak negative correlation was found
etween MSG and GR Topt (r2 = 0.46, P = 0.0071). b

Switchgrass (Panicum virgatum L.) Intraspecific Variation and Thermotolerance Classification 142
Using in Vitro Seed Germination Assay
Table 4. Classification of switchgrass genotypes into (a) heat-tolerance and (b) cold-tolerance groups based on cumulative
temperature response index (CTRI; unitless) along with individual scores in parenthesis.
(a) Heat-tolerance classification based on CTRI
Heat-sensitive
(CTRI = 4.83 - 5.43)
Intermediate
(CTRI = 5.44 -5.74)
Heat-tolerant
(CTRI = 5.75 - 6.05)
Expresso (4.83) Alamo (5.45) Summer (5.78)
Trailblazer (4.85) Blackwell (5.47)
Sunburst (5.0) Shawnee (5.51)
Cave-in-Rock (5.01) Carthage (5.56)
Kanlow (5.03) Shelter (5.59)
Tusca (5.06)
Forestburg (5.16)
Dacotah (5.36)
(b) Cold-tolerance classification based on CTRI
Cold-sensitive
(CTRI = 4.74 - 5.03)
Moderately cold-sensitive
(CTRI = 5.04 - 5.32)
Moderately cold-tolerant
(CTRI = 5.33 -5.62)
Cold-tolerant
(CTRI = 5.63 - 6.21)
Shelter (4.74) Forestburg, (5.08) Trailblazer (5.52) Expresso (5.64)
Summer (4.74) Tusca (5.19)
Carthage (4.78) Kanlow (5.21)
Shawnee (4.8) Cave-in-Rock (5.24)
Blackwell (4.82) Sunburst (5.26)
Alamo (4.84)
Dacotah (5.0)
4. Discussion
Seed germination is a complex physiological process
modulated by internal and external factors and their in-
teractions. Similar to other growth and developmental
processes, temperature influences seed dormancy, ger-
mination capacity and rate, and seedling emergence. To
our knowledge, this is the first study to evaluate the in-
fluence of temperature effects on seed germination char-
acteristics of diverse switchgrass genotypes. The result-
ing data provided functional algorithms for modeling and
segregating genotypes for cold- and heat-tolerance based
on seed-based parameters.
Optimal temperatures for MSG and GR differed
among the genotypes with MSG optimum occurring over
a range and GR having a sharply defined optimum. Rela-
tive to MSG, GR had higher Tmin, Topt and Tmax values
consistent with previous reports that many species typi-
cally have higher optimum temperatures for GR than for
MSG percentage [11]. Germination rate is more tem-
perature sensitive than final germination percentage in
Setaria lutescens and Amaranthus retroflexus [37] simi-
lar to our finding in switchgrass genotypes. Germination
rate is affected by the depth of dormancy, imbibition rate
and the rate of catabolic and anabolic pathways all of
which are directly or indirectly temperature dependent
while the maximum seed germination is more affected by
the rate of rehydration rather than the speed of the
physiological pathways affecting cell expansion.
4.1. Maximum Seed Germination
All switchgrass genotypes tested exhibited a quadratic
response to temperature (r2 = 0.93), similar to indian-
grass (Sorghastrum nutans (L.) Nash) tested under alter-
nating temperature conditions [6], another native
warm-season species. Mean MSG (73%) in the current
study is similar to the 77% to 78% reported for similar
genotypes [1,2,5], although the temperature and lighting
conditions across these experiments are divergent.
With the exception of Expresso, which has been se-
lected for increased precocious germination (B. Baldwin,
personal communication, 2009), MSG of the other two
lowland genotypes (Alamo and Kanlow) were less than
55%.
The optimum temperature for switchgrass MSG in the

Switchgrass (Panicum virgatum L.) Intraspecific Variation and Thermotolerance Classification 143
Using in Vitro Seed Germination Assay
current study varied between 24.04˚C and 28.56˚C
among the genotypes, which is within the range of values
reported in other warm-season grasses; 20˚C to 30˚C for
Cane beardgrass [Bothriochloa barbinodis (Lag.) Herter],
sideoats grama [Bouteloua curtipendula (Michx.) Torr.],
and tanglehead [Heteropogon contortus (L.) P. Beauv. ex
Roem. & Schult.] [38] and 16.5˚C to 27˚C for indian-
grass [39]. Maximum seed germination minimum tem-
perature averaged 8.08˚C and ranged from 3.69 to
12.83˚C, which is similar to Tmin of other warm-season
grasses [40]; 5.5˚C to 10.9˚C for switchgrass, 7.3˚C to
8.7˚C for big bluestem (Andropogon gerardii Vitman),
7.5˚C to 9.6˚C for indiangrass, and 4.5˚C to 7.9˚C for
prairie sandreed (Calamovilfa longifolia (Hook.) Scribn.).
4.2. Germination Rate
Thermal response of switchgrass seed germination is
consistent with thermal response patterns of a number of
other physiological processes [41]. At suboptimal tem-
peratures (Tmin to Topt), germination rate (reciprocal time
to 50% germination) generally increases linearly with
temperature, but decreases linearly with temperature at
supra-optimal temperatures (Topt to Tmax). This character-
istic thermal response is similar to germination rate of
chickpea (Cicer arietinum L.) [2,27], lentil (Lens culi-
naris Medic.) and soybean (Glycine max (L.) Merr.) [27],
pearl millet (Pennisetum glaucum (L.) R. Br.) [42], sor-
ghum [Sorghum bicolor (L.) Moench.] [43] and cool
season weeds [44]. A decline in germination rate with
decreasing temperature is partly associated with decline
in the imbibition rate observed with a reduction in tem-
perature [45]. Germination rate response to temperature
was described previously by two linear equations; the
first describing the positive linear relationship between
the minimum and optimum temperatures and the second
describing the negative linear relationship between opti-
mum and maximum temperature [2,27]. In the current
study, GR was modeled using a single modified bilinear
equation, which was previously used by several studies
([30-33,46]) to quantify pollen germination and pollen
tube growth responses to temperature. Analogous to pol-
len, seeds are considered independent functional units
that are responsive to temperature changes.
Even though MSG percentage is the most important
parameter determining commercial value of seedlots, GR
influences the uniformity and rapidity of emergence in
nurseries [47]. Germination rates are most rapid at opti-
mum temperature ranging from 29.5˚C to 35.6˚C.
4.3. Cardinal Temperatures
Biological processes are typically characterized by car-
dinal temperatures describing the range of temperature
over which a process can occur. The effect of tempera-
ture on seed germination can be expressed in terms of car-
dinal temperatures, that is, Tmin, Topt, and Tmax at which
germination will occur [48]. Cardinal temperatures may
be used to describe the range of adaptation of a species.
Though switchgrass is reported to be the most tem-
perature specific of the warm-season grasses [15], there
exists significant intra-specific differences in cardinal
temperatures that may be related to the different areas of
origin or adaptation [40,49]. The genotypes Cave-in-
Rock, Dacotah, Forestburg, Shawnee, Shelter, Summer,
Sunburst, and Trailblazer are from the cooler northern
regions where average minimum temperatures range
from –23.3˚C to –17.8˚C, while Alamo, Blackwell, Ex-
presso, Kanlow, and Tusca are from warmer growing
regions with average minimum temperatures ranging
from –17.8˚C to 4.4˚C. Cardinal temperature coefficients
can be directly compared for screening germplasm [44].
The cardinal temperatures derived for both MSG and GR
can be used in evaluation of potential regions for intro-
duction of switchgrass and also aid in on-farm opera-
tional practices such as appropriate sowing dates when
soil temperature would be conducive to optimum germi-
nation and emergence and ultimately optimum stand es-
tablishment and crop performance. Genotypes with lower
Tmin values can be subjected to early-season sowing be-
cause of their inherent capacity to germinate in cooler
temperatures. The variability of cardinal temperatures
both for MSG and GR indicates broad latitudinal adapta-
tion across the various plant hardiness zones of the USA
[50].
The cardinal temperatures derived for GR may be
comparable with subsequent developmental stages of
switchgrass ontogeny (morphological development). Kiniry
et al. [16] assumed a base temperature of 12˚C for all
growth stages of switchgrass in the ALMANAC model,
however, the results in this study suggest that cardinal
temperatures are genotype-specific and may be proc-
ess-specific as well. Therefore, the derived cardinal tem-
peratures in this study may be used to refine model algo-
rithms for on-farm application and policy assessments.
4.4. Temperature Tolerance Classification
Temperature tolerance refers to the ability of an organ-
ism to cope with excessively high or low temperatures.
Direct selection under field conditions is generally diffi-
cult because uncontrollable environmental factors affect
the precision and repeatability of such trials. Stress tol-
erance is a developmentally regulated, stage-specific
phenomenon; hence species may show different sensitiv-
ity to stress at different developmental stages. All stages
through a plant’s ontogeny are sensitive to temperature;

Switchgrass (Panicum virgatum L.) Intraspecific Variation and Thermotolerance Classification
Using in Vitro Seed Germination Assay
144
will validate the use of seed-based parameters as a
screening tool. This information is lacking in the litera-
ture with respect to screening temperature tolerance of
diverse switchgrass genotypes, even though several
studies link intraspecific differences in germination to
geographical and ecological areas of distribution or ori-
gin [51].
therefore, screening for tolerance should be conducted at
the most sensitive stage. Seed germination is temperature
dependent and can be used to screen for temperature tol-
erance. In vitro assays are not subjected to uncontrollable
biotic and abiotic stress factors marring true tolerance
potential. In the field, genotypes with high minimum
temperature would experience little germination in early
spring when temperatures would frequently drop below
the Tmin level.
The classification method tested suggests that CTRI
for heat- and cold-tolerance are inversely related (r2 =
0.64, P = 0.0006), suggesting that it may be difficult to
identify a cultivar that possesses both heat- and cold-
tolerance characteristics (Figure 4). Variability among
genotypes for heat- and cold-tolerance suggests that se-
lection or breeding among genotypes is a viable objective.
Switchgrass adaptation to a specific ecoclimatic and ed-
aphic region is determined by the growth rate, photope-
riodism, heat tolerance, and cold or freezing tolerance of
a specific genotype [10].
In the current study, the successful use of CTRI, based
on the summation of individual temperature response
indices and then separated by standard deviation based
on the number of classes of interest, confirms that
seed-based parameters derived from in vitro seed germi-
nation assay can be used for genotype temperature toler-
ance classification. Genotype variability associated with
temperature tolerance was demonstrated in this study.
Alamo, Blackwell, Carthage, Dacotah, Shawnee, Shelter,
and Summer were classified as cold-sensitive while Ex-
presso was classified as cold-tolerant. Conversely, Cave-
in-Rock, Dacotah, Expresso, Forestburg, Kanlow, Sun-
burst, Trailblazer, and Tusca were determined to be
heat-sensitive and Summer as heat-tolerant. Since basal
temperature tolerance is a function of genetics and ac-
quired temperature tolerance is latitude and tempera-
ture-induced, corroborating seed-based temperature tol-
erance with vegetative or other reproductive responses
Ecotype classification in this study did not necessarily
confer the temperature tolerance characteristic of a spe-
cific ecotype. For example, Alamo, a lowland genotype,
was classified as intermediately heat-tolerant while
Summer, an upland genotype was classified as heat-
tolerant using seed-based parameters. Genotype tempera-
ture tolerance is determined not only by ecotypic classi-
fication, but also latitude of origin, photoperiodism and
Figure 4. The relationship between heat- and cold-tolerance cumulative temperature response index (CTRI) for 14 swi tchgr ass
genotypes.

Switchgrass (Panicum virgatum L.) Intraspecific Variation and Thermotolerance Classification 145
Using in Vitro Seed Germination Assay
genetics. Being photoperiod sensitive [52], switchgrass
morphological development is determined primarily by
its response to daylength [53]. Since ecotypic classifica-
tion are more related to photoperiod responsiveness than
temperature, the small or little variation observed be-
tween upland and lowland ecotypes for seed germination
characteristics may be as result of ecotypic temperature
insensitivity.
Since tolerance mechanisms are developmentally
regulated, it is prudent to validate controlled in vitro seed
germination assay with field performance tests. In the
current study, GR and MSG were evaluated as estimators
of temperature tolerance using 14 diverse genotypes.
Using similar techniques, 12 genotypes of sorghum were
screened for cold tolerance in controlled in vitro germi-
nation studies and GR was found to be strongly corre-
lated with rate of emergence under field conditions, con-
firming that screening using parameters based on in vitro
studies is a rapid and reliable method for handling large
number of genotypes before evaluation in the field [26].
The current study quantified the relation between GR and
temperature, highlighting genotypic differences. It is
necessary in future work, therefore, to determine whether
in vitro seed germination assay has potential in selection
and screening procedures in breeding programs.
5. Conclusions
The current study quantified the effects of temperature
on seed germination rate and capacity of 14 diverse
switchgrass genotypes and determined the cardinal tem-
peratures for MSG and GR. Genotypic variability for
MSG, GR, their respective cardinal temperatures, and
TAR were found to exist among the switchgrass geno-
types tested. Mean minimum temperatures for MSG and
GR were 8.08˚C and 11.1˚C, respectively, while opti-
mum temperatures were 26.6˚C and 33.1˚C, respectively.
The cumulative temperature response index method used
in the current study identified both heat and cold tolerant
genotypes and demonstrated that variability existed
among genotypes and ecotypes. The cardinal temperature
estimates would be useful to improve switchgrass models
for field applications. Additionally, the identified cold-
and heat-tolerant genotypes can be selected for niche
environments and in switchgrass breeding programs to
develop new genotypes for cold and hot environments.
6. Acknowledgements
The authors are grateful to Ernst Seed Company for
providing seeds of 5 switchgrass genotypes. This research
was funded in part by the Department of Energy through
Sustainable Energy Center, Mississippi State University,
Mississippi State, MS, the USDA-UV-B Monitoring and
Research Program, and the USDA-ARS 58-6402-7-241.
This article has been approved for publication as Journal
Article No. J11898 of the Mississippi Agricultural and
Forestry Experiment Station, Mississippi State University.
REFERENCES
[1] G. Hacisalihoglu, “Responses of Three Switchgrass
(Panicum virgatum L.) Cultivars to Seed Priming and
Differential Aging Conditions,” Acta Agriculturae Scan-
dinavica Section B-Soil and Plant Science, Vol. 58, No. 3,
2008, pp. 280-284.
[2] G. E. Aiken and T. L. Springer, “Seed Size Distribution,
Germination, and Emergence of Six Switchgrass Culti-
vars,” Journal of Range Ma nagement, Vol. 48, No. 5, 1995,
pp. 455-458. doi:10.2307/4002252
[3] F. H. Hsu, C. J. Nelson and W. S. Chow, “A Mathemati-
cal Model to Utilize the Logistic Function in Germination
and Seedling Growth,” Journal of Experimental Botany,
Vol. 35, No. 11, 1984, pp. 1629-1640.
doi:10.1093/jxb/35.11.1629
[4] W. Robocker, J. Curtis and H. Ahlgren, “Some Factors
Affecting Emergence and Establishment of Native Grass
Seedlings in Wisconsin,” Ecology, Vol. 34, No. 1, 1953,
pp. 194-199. doi:10.2307/1930322
[5] J. D. Hanson and H. A. Johnson, “Germination of Swi-
tchgrass under Various Temperature and pH Regimes,”
Seed Technology, Vol. 27, No. 2, 2005, pp. 203-210.
[6] T. E. Fulbright, “Effects of Temperature, Water Potential
and Sodium Chloride on Indiangrass Germination,” Jour-
nal of Range Management, Vol. 41, No. 3, 1988, pp. 207-
210. doi:10.2307/3899168
[7] S. B. McLaughlin and M. E. Walsh, “Evaluating Envi-
ronmental Consequences of Producing Herbaceous Crops
for Bioenergy,” Biomass and Bioenergy, Vol. 14, No. 4,
1998, pp. 317-324.
doi:10.1016/S0961-9534(97)10066-6
[8] M. K. Das, R. G. Fuentes and C. M. Taliaferro, “Genetic
Variability and Trait Relationships in Switchgrass,” Crop
Science, Vol. 44, No. 2, 2004, pp. 443-448.
[9] S. A. Eberhart and L. C. Newell, “Variation in Domestic
Collections of Switchgrass, Panicum virgatum L,”
Agronomy Journal, Vol. 51, No. 10, 1959, pp. 613-616.
doi:10.2134/agronj1959.00021962005100100015x
[10] M. D. Casler, et al., “Latitudinal and Longitudinal Adap-
tation of Switchgrass Populations,” Crop Science, Vol. 47,
No. 6, 2007, pp. 2249-2260.
doi:10.2135/cropsci2006.12.0780
[11] E. H. Roberts, “Temperature and Seed Germination,” In:
S. P. Long and F. I. Woodword, Eds., Plants and Tem-
perature, Symposia of the Society for Experimental Biol-
ogy, Cambridge, 1988, pp. 109-132.
[12] G. L. Jordan and M. R. Haferkamp, “Temperature Re-
sponses and Calculated Heat Units for Germination of
Several Range Grasses and Shrubs,” Journal of Range

Switchgrass (Panicum virgatum L.) Intraspecific Variation and Thermotolerance Classification 146
Using in Vitro Seed Germination Assay
Management, Vol. 42, No. 1, 1989, pp. 41-45.
doi:10.2307/3899656
[13] R. Ellis, E. Roberts and P. Hebblethwaite, “Towards a
Rational Basis for Testing Seed Quality,” In: P. Heb-
blethwaite, Ed., Seed Production, Butterworths, London,
1980, pp. 605-635.
[14] D. J. Parrish and J. H. Fike, “The Biology and Agronomy
of Switchgrass for Biofuels,” Critical Reviews in Plant
Sciences, Vol. 24, No. 5-6, 2005, pp. 423-459.
doi:10.1080/07352680500316433
[15] F. H. Hsu, C. J. Nelson and A. G. Matches, “Temperature
Effects on Germination of Perennial Warm-Season For-
age Grasses,” Crop Science, Vol. 25, No. 2, 1985, pp.
215-215.
doi:10.2135/cropsci1985.0011183X002500020005x
[16] J. R. Kiniry, et al., “Switchgrass Simulation by the AL-
MANAC Model at Diverse Sites in the Southern US,”
Biomass and Bioenergy, Vol. 29, No. 6, 2005, pp. 419-425.
doi:10.1016/j.biombioe.2005.06.003
[17] E. Heaton, T. Voigt and S. P. Long, “A Quantitative Re-
view Comparing the Yields of Two Candidate C4 Peren-
nial Biomass Crops in Relation to Nitrogen, Temperature
and Water,” Biomass and Bioenergy, Vol. 27, No. 1, 2004,
pp. 21-30. doi:10.1016/j.biombioe.2003.10.005
[18] S. McLaughlin, et al., “Developing Switchgrass as a Bio-
energy Crop,” In: J. Janick, Ed., Perspectives on New
Crops and New Uses, ASHS Press, Alexandria, 1999. pp.
282-289.
[19] P. S. Setimela, et al., “Screening Sorghum Seedlings for
Heat Tolerance Using a Laboratory Method,” European
Journal of Agronomy, Vol. 23, No. 2, 2005, pp. 103-107.
doi:10.1016/j.eja.2004.09.008
[20] M. Bouslama and W. T. Schapaugh Jr, “Stress Tolerance
in Soybeans. I. Evaluation of Three Screening Techniques
for Heat and Drought Tolerance,” Crop Science, Vol. 24,
No. 5, 1984, pp. 933-933.
doi:10.2135/cropsci1984.0011183X002400050026x
[21] S. Sadasivam, et al., “Genetic Variation in Seed Germi-
nation, Root Traits and Drought Recovery in Rice,” In-
dian Journal of Plant Physiology, Vol. 5, No. 1, 2000, pp.
73-78.
[22] M. R. Foolad and G. Y. Lin, “Genetic Potential for Salt
Tolerance during Germination in Lycopersicon Species,”
HortScience, Vol. 32, 1997, pp. 296-300.
[23] N. Misra and U. N. Dwivedi, “Genotypic Difference in
Salinity Tolerance of Green Gram Cultivars,” Plant Sci-
ence, Vol. 166, No. 5, 2004, pp. 1135-1142.
doi:10.1016/j.plantsci.2003.11.028
[24] F. F. Hou and F. S. Thseng, “Studies on the Screening
Technique for Pre-Germination Flooding Tolerance in
Soybean,” Japanese Journal of Crop Science, Vol. 61,
No. 3, 1992, pp. 447-453.
doi:10.1016/j.plantsci.2003.11.028
[25] S. N. Acharya, J. Dueck and R. K. Downey, “Selection
and Heritability Studies on Canola/Rapeseed for Low
Temperature Germination,” Canadian Journal of Plant
Science, Vol. 63, No. 2, 1983, pp. 377-384.
doi:10.4141/cjps83-043
[26] I. Tiryaki and D. J. Andrews, “Germination and Seedling
Cold Tolerance in Sorghum. I. Evaluation of Rapid
Screening Methods,” Agronomy Journal, Vol. 93, No. 6,
2001, pp. 1386-1391. doi:10.2134/agronj2001.1386
[27] S. Covell, et al., “The Influence of Temperature on Seed
Germination Rate in Grain Legumes. I. A Comparison of
Chickpea, Lentil, Soybean and Cowpea at Constant
Temperatures,” Journal of Experimental Botany, Vol. 37,
No. 5, 1986, pp. 705-715. doi:10.1093/jxb/37.5.705
[28] B. N. Emerson and H. C. MiNor, “Response of Soybeans
to High Temperature during Germination,” Crop Science,
Vol. 19, No. 4, 1979, pp. 553-553.
doi:10.2135/cropsci1979.0011183X001900040029x
[29] S. Koti, et al., “Soybean (Glycine Max) Pollen Germina-
tion Characteristics, Flower and Pollen Morphology in
Response to Enhanced Ultraviolet-B Radiation,” Annals
of Botany, Vol. 94, No. 6, 2004, pp. 855.
doi:10.1093/aob/mch212
[30] M. A. Salem, et al., “Pollen-based Screening of Soybean
Genotypes for High Temperatures,” Crop Science, Vol.
47, No. 1, 2007, pp. 219-231.
doi:10.2135/cropsci2006.07.0443
[31] V. G. Kakani, et al., “Response of in Vitro Pollen Ger-
mination and Pollen Tube Growth of Groundnut (Arachis
hypogaea L.) Genotypes to Temperature,” Plant Cell and
Environment, Vol. 25, No. 12, 2002, pp. 1651-1661.
doi:10.1046/j.1365-3040.2002.00943.x
[32] V. G. Kakani, et al., “Differences in in vitro Pollen Ger-
mination and Pollen Tube Growth of Cotton Cultivars in
Response to High Temperature,” Annals of Botany, Vol.
96, No. 1, 2005, pp. 59-67. doi:10.1093/aob/mci149
[33] S. K. Singh, et al., “Assessment of Cold and Heat Toler-
ance of Winter-Grown Canola (Brassica napus L.) Culti-
vars by Pollen-Based Parameters,” Journal of Agronomy
and Crop Science, Vol. 194, No. 3, 2008, pp. 225-236.
doi:10.1111/j.1439-037X.2008.00309.x
[34] Z. X. Shen, et al., “Stratification in Switchgrass Seeds is
Reversed and Hastened by Drying,” Crop Science, Vol.
41, No. 5, 2001, pp. 1546-1551.
doi:10.2135/cropsci2001.4151546x
[35] Systat Software Inc., “SigmaPlot 10.0—User’s Guide,”
Point Richmond, 2006.
[36] SAS Institute Inc., “SAS/STAT User’s Guide, Version
9.1.3.” SAS Institute Inc., Cary, 2004.
[37] D. J. Schimpf, S. D. Flint and I. G. Palmblad, “Represen-
tation of Germination Curves with the Logistic Function,”
Annals of Botany, Vol. 41, No. 6, 1977, pp. 1357-1360.
[38] B. A. Roundy and S. H. Biedenbender, “Germination of
Warm-Season Grasses under Constant and Dynamic
Temperatures,” Journal of Range Management, Vol. 49,
No. 5, 1996, pp. 425-431. doi:10.2307/4002924
[39] D. G. Sabo and R. M. Forest, “Germination Requirements
of 19 Species of Arid Land Plants,” Rocky Mountain

Switchgrass (Panicum virgatum L.) Intraspecific Variation and Thermotolerance Classification
Using in Vitro Seed Germination Assay
147
Forest and Range Experiment Station, Forest Service, US
Department of Agriculture, 1979.
[40] I. C Madakadze, et al., “Variation in Base Temperatures
for Germination in Warm Season Grasses,” Seed Science
and Technology, Vol. 29, No. 1, 2001, pp. 31-38.
[41] R. Probert, “The Role of Temperature in the Regulation
of Seed Dormancy and Germination,” In: M. Fenner, Ed.,
The Ecology of Regeneration in Plant Communities,
CABI Publishing, Wallingford, 2000, pp. 261-292.
doi:10.1079/9780851994321.0261
[42] J. Garcia-Huidobro, J. L. Monteith and G. R. Squire,
“Time, Temperature and Germination of Pearl Millet
(Pennisetum Typhoides S. & H.) I. Constant Tempera-
ture,” Journal of Experimental Botany, Vol. 33, No. 2,
1982, pp. 288-296. doi:10.1093/jxb/33.2.288
[43] R. L. Benech-ArNold, et al., “Temperature Effects on
Dormancy Release and Germination Rate in Sorghum
halepense (L.) Pers. Seeds: A Quantitative Analysis,”
Weed Research, Vol. 30, No. 2, 1990, pp. 81 - 89.
[44] S. P. Hardegree, “Predicting Germination Response to
Temperature. I. Cardinal-Temperature Models and Sub-
population-Specific Regression,” Annals of Botany, Vol.
97, No. 6, 2006, pp. 1115-1125. doi:10.1093/aob/mcl071
[45] M. Lopez, et al., “The Effect of Temperature and Water
Stress on Laboratory Germination of Eucalyptus Globu-
lus Labill. Seeds of Different Sizes,” Annals of Forest
Science, Vol. 57, No. 3, 2000, pp. 245-250.
doi:10.1051/forest:2000115
[46] K. R. Reddy and V. G. Kakani, “Screening Capsicum
Species of Different Origins for High Temperature Tol-
erance by in Vitro Pollen Germination and Pollen Tube
Length,” Scientia Horticulturae, Vol. 112, No. 2, 2007, pp.
130-135. doi:10.1016/j.scienta.2006.12.014
[47] Y. A. El-Kassaby, et al., “Seed Germination: Mathemati-
cal Representation and Parameters Extraction,” Forest
Science, Vol. 54, No. 2, 2008, pp. 220-227.
[48] L. Copeland and M. B. McDonald, “Principles of Seed
Science and Technology,” 4th Edtion, Springer, Berlin,
2001.
[49] M. D. Casler and A. R. Boe, “Cultivar X Environment
Interactions in Switchgrass,” Crop Science, Vol. 43, No.
6, 2003, pp. 2226-2233. doi:10.2135/cropsci2003.2226
[50] M. D. Casler, et al., “Latitudinal Adaptation of Switch-
grass Populations,” Crop Science, Vol. 44, No. 1, 2004,
pp. 293-303.
[51] A. Orozco-Segovia, et al., “A Mathematical Model that
Uses Gaussian Distribution to Analyze the Germination
of Manfreda brachystachya (Agavaceae) in a Thermo-
gradient,” Physiologia Plantarum, Vol. 98, No. 3, 1996,
pp. 431-438. doi:10.1111/j.1399-3054.1996.tb05696.x
[52] L. E. Moser and K. P. Vogel, “Switchgrass, Big Bluestem
and Indiangrass,” In: R. F. Barnes, D. A. Miller and C. J.
Nelson, Eds., An Introduction to Grassland Agriculture,
Iowa State University Press, Ames, 1995, pp. 409-420.
[53] R. Mitchell and L. Moser, “Developmental Morphology
and Tiller Dynamics of Warm-Season Grass Swards,” In:
Anderson, Ed., CSSA, Madison, 2000, pp. 49-66.