The Influence of Temperature on Seed Germination in Cultivars of Common Bean

Author(s): JEFFREY W. WHITE and CONSUELO MONTES-R

Source: Journal of Experimental Botany , December 1993, Vol. 44, No. 269 (December
1993), pp. 1795-1800
Published by: Oxford University Press

Stable URL: https://www.jstor.org/stable/23694697

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Oxford University Press is collaborating with JSTOR to digitize, preserve and extend access to
Journal of Experimental Botany

This content downloaded from

132.174.251.190 on Thu, 07 Sep 2023 02:55:56 +00:00
All use subject to https://about.jstor.org/terms
Journal of Experimental Botany, Vol. 44, No. 269, pp. 1795-1800, December 1993

The Influence of Temperature on Seed

Germination in Cultivars of Common Bean

JEFFREY W. WHITE1 and CONSUELO MONTES-R

Bean Physiology, CIAT, Apartado Aereo 6713, Cali, Colombia

Received 5 July 1993; Accepted 13 September 1993

ABSTRACT
Common beau (Phaseolus vulgaris L.) is grown over a wide range of environments, including sit
temperatures at sowing time. To describe the temperature response of seed germination, 20 bean
rolled paper towel system with 11 constant temperatures ranging from 12 to 34 °C. Germination
fitting cumulative counts using a maximum-likelihood analysis. Rate of germination increased fro
typically near 8 °C to an optimal development temperature (T0) of 29 to 34 °C. Tb did not differ
Mesoamerican germplasm showed slightly higher Ta than Andean germplasm, but there was lar
the two gene pools. The single accession of tepary bean (P. acutifolius) evaluated appeared to be
germination temperatures.

Key words: Common bean, seed germination rate, temperature.

INTRODUCTION

The temperature response of seed germination in common response to germinat

bean is of interest for many reasons. Beans are grown in origin are also of intere
a wide range of environments (Laing et al., 1984). At Although there have been
higher latitudes and highland tropical sites, low temper- response of germinat
atures frequently cause slow germination and lead to Waines, 1987; Hoogen
poor stand establishment. Selection for genotypes show- have often used t
ing improved germination at low temperatures was effect- temperatures to quant
ive in improving crop performance in cool environments Furthermore, some stu
(Dickson, 1971; Hardwick and Andrews, 1980). Models istical models. Germinatio
which account for genotypic effects on crop growth and since they are cumul
production usually predict phenology and, unless the correlated, and have
germination phase is ignored, this requires an understand- distribution (Hunter
ing of temperature effects on germination. The model for analysed using conve
dry bean, Beangro (Version 101), estimates the time from (Brain and Butler, 198
planting to seedling emergence using only air temperatures marize data only usi
and assumes that the temperature response of germination mination or time to a
is the same as for other processes of vegetative develop- the possible effects of
ment (Hoogenboom et al., 1991). However, this approach tion (Scott et al., 1984
has not been evaluated rigorously. Recent descriptions of One approach for par
gene pools and races in common bean have characterized problems is to assume t
the Andean gene pool as being more adapted to cooler temperature have a m
environments than the Mesoamerican one (Singh, 1989; et al., 1984; Brain and B
Singh et al., 1991), so comparisons of temperature mination to be estimated us

1 To whom correspondence should be addressed.

© Oxford University Press 1993

This content downloaded from

132.174.251.190 on Thu, 07 Sep 2023 02:55:56 +00:00
All use subject to https://about.jstor.org/terms
1796 White and Montes-R—Seed Germination of Common Bean
function In an initial trial to determine an appropriate range of temper
atures, the common bean genotypes BAT 477
G(t) =N(b[F(t — l)—m]), 'Diacol Calima' (CIAT accession G 4494), and the tepary bean
accession Sonora 32 (G 40159A) were evaluated. In a second
where N(x) is the normal probability integral, t is time, trial, 16 genotypes were evaluated. Seed size, origin and bean
/ is the lag time before onset of germination events, b is race of all materials are listed in Table 1. Races Durango and
the reciprocal of the variance of F(t-l) and m is the Mesoamerican belong to the Mesoamerican gene pool and race
mean of F(t-l). F(x) is a function to transform the time Nueva Granada belongs to the Andean gene pool (Singh
• i , 1/y 1 )<
variable. Seeds were incubated on germination paper (No. 76 from
With estimates of G(t) for various temperature treat- Anchor Paper, St Paul, MN, USA
ments, the effects of temperature on mean time to ger- 254 mm. Twenty-five seeds per roll wer
mination or other parameters are easily examined. Rate treated with a mixture of benomyl (meth
of progress of cumulative germination, defined as the benzimidazol-2-ili-carbonate), carboxi
r . . i , ✓ ™«/ • iV-phenyl-l,4-oxatiin-3-carboxamide) and captan (as-AHtnchl
inverse of mean time to a given level (e.g. 50/o germina- oromethyl)-thio-4-cyclohexane-l,2
tion), often shows a simple response pattern such that and bacteria. Seed rolls were p
rate increases linearly up to the optimal temperature and measuring 103 mm high by
decreases linearly above this temperature (Garcia- tight-fitting lids. Eight rolls per
Huidobro et. . al.,
. . ' 1982; Covell
intervals, et ai, every
initially 1986; 5
Ellis
to 6et
h al,
once. tProSress t°w1fds
germination began. §er™"at^n
Seeds wa
1986). Since estimation of G(t) requires an estimate of were considered to have ger
maximum germination for a given temperature, temper- least 10 mm long. Time to
ature effects on maximum germination can be examined roots was also recorded, b
along with other parameters response appeared similar to those
-r, , • .. c . , . ,, • for secondary roots are not presented. Rolls were remoistened
The objective of our study was to examine the germina- , , 7 F
J J 6 as needed with distilled water, and seeds were discarded after
tion response to temperature for contrasting bean geno- secondary roots
types. By analysing data using the cumulative distribution In the first experim
function G(t), statistical problems related to serial cor- to 34-9 °C were o
relation and lack of normality of count data were avoided. constructed f
Modifications included the use of a conventional incubator

uiTUBTAi q a \in ucTHOnc cooling system plus resistance heaters to create the temperature
AINU MfcinuiJb gradient and increasing the size of the aluminium cabinet t
Seeds were obtained from field plots at the Palmira station of 1-2 m tall by 0-3 m by 0-3 m, permitting a vertical stack of
CIAT, which were managed to avoid nutrient and water deficits. chambers. A 12 °C treatment was obtained using a conventio

Table 1. Characteristics of the common bean genotypes evaluated in this investigation

Identification CIAT No." Origin1' Seed weight Growth" Raced
(mg) habit

A 22 220 III Mesoamerican

A 36 390 III Nueva Granada
A 195 540 I Nueva Granada
A 197 490 I Nueva Granada
BAT 477 220 III Mesoamerican
BAT 881 180 II Mesoamerican
BAT 1297 220 III Mesoamerican
BAT 1393 360 I Nueva Granada
Carioca G 4017 BZL 240 III Mesoamerican
Diacol Calima G 4494 CLB 480 I Nueva Granada
ICA Linea 24 G 13922 CLB 490 Nueva Granada
Jamapa G 3645 MEX 200 Mesoamerican
Magdalena 3 G 2525 CLB 180 Mesoamerican
NEP Bayo 22 G 4000 CRA 190 Mesoamerican
Pinto UI 114 G 4449 USA 360 Durango
Porrillo Sintetico G 4495 ELS 210 II Mesoamerican
San Cristobal 83 G 17722 RDM 280 III Mesoamerican
WAF 10 290 I Nueva Granada
XAN 90 250 Mesoamerican
Sonora 32 G 40159A MEX 120 III

" Accession number for collection held by Genetic Resources Unit at CIAT.
b Genotypes with no origin listed are CIAT breeding lines.
"By CIAT standard evaluation system (Schoonhoven and Pastor Corrales, 1987) where I = determinate bush,
II = indeterminate, erect bush, III = indeterminate, prostrate bush, and IV = indeterminate climber.
d Races are as defined by Singh et al. (1991), except for the P. acutifolius accession Sonora 32 since races have
not been defined for this species.

This content downloaded from

132.174.251.190 on Thu, 07 Sep 2023 02:55:56 +00:00
All use subject to https://about.jstor.org/terms
 White and Montes-R—Se
constant temperature incubator. In the 16 geno
temperatures in the gradient incubator were from
and the 12 °C incubator was also used. Temper
were not used since preliminary trials showed
posed at higher temperatures regardless of se
control pathogens.
Two boxes of seeds per temperature treatm
Four genotypes were evaluated at one time (f
seeds per genotype). In the first experiment, th
considered as replicates. In the second study, da
were combined and separate runs were used as
Germination data were analysed by fitting c
normal distribution assuming a Poisson error
permitting a lag period before the onset of ger
et al., 1984; Brain and Butler 1988). Maximum
assumed equal to the highest value obtained for
The rate of progress of cumulative germinat
referred to as 'rate of germination', was c
reciprocal of mean time to germination (expre
including the lag phase) multiplied by 100 to pe
tion of the rate in terms of per cent complet
This method of calculation differs slightly fr
Garcia-Huidobro et al. (1982), since they used tim
mined level of germination. The relation betw
mination and temperature was described using
to those of Garcia-Huidobro et al. (1982) an
(1986), where rate increases linearly with tempe
optimal temperature. By fitting a segmented l
base (Tb) and optimal (Ta) development tem
maximum rate of germination were estimated.
50
50 75 75 100
above T0, the rate was assumed to continue at
Time from onset (h)
value since insufficient data wer
temperature (Tc of Garcia-Huidobro
Fig. et al.
1. Seed germination o
Germination curves were fitted
and 32-5 using
°C (•). Gensta
Points are
Committee, 1987). Subsequent analyses
fitted curves, were
(a) BAT 477c
the SAS package (SAS Institute, 1985).

RESULTS AND DISCUSSION 100100

The model for seed germination proposed by Hunter

et al. (1984) was found to describe the data satisfactorily
c 95
if the time-scale was transformed
0 using the square roo
"co
function of time rather than with the logarithmic trans- c

formation they used. Examples 1¡59090of curves estimated for

D)
O)
the model are given for BAT E 477 and Sonora 32 germin
D 13
ated at 16-2 °C and 32-5 °C (Fig.
E 1). Maximum germina
>< 85
tion was affected far more by extreme temperatures for
Diacol Calima than for the other genotypes (Fig.
BAT 477 BAT 1297 Diacol Calima Sonora 32 2).
A •
+ --43
---B--
— A
A
In the first trial, rate of germination 80 increased linearl
up to values of T0 from 27-0 °C to 34-5 °C (Fig. 3;
Table 2). At higher temperatures, rate of germination 10 15 20 25 30 35
10 15 20 25 30 35

appeared to continue at the maximal value or decline Temperature (°C)

Temperature (°C

slightly. However, a precise relation was not determined Flo 2 Effect

Fig. of temperature
2. Effect on per cent maximum observed germina
of tempera
due to the narrow range of temperatures where seed tion fortion four bean genotypes.
for Data fourare meansbean
of two replicates,
genoty
showed a constant or reduced rate of germination and
did not decompose rapidly, as occurred at temperatures The most notable difference among the four genotypes
above 35 °C in preliminary trials (data not shown). A was in maximum rate of germination, with Diacol Calima
similar problem in estimating a maximum was mentioned having the lowest rate and Sonora 32 the highest
by Ellis et al. (1986) in their comparison of chickpea (Table 2). Based on comparisons using approximate 95%
(Cicer arietinum L.) genotypes, and is common in studies confidence intervals calculated from standard errors,
of temperature effects on crop development (Ritchie and values of Tb did not differ, but Sonora 32 had a higher
Ne Smith, 1991). Ta than the P. vulgaris genotypes. This agreed with

This content downloaded from

132.174.251.190 on Thu, 07 Sep 2023 02:55:56 +00:00
All use subject to https://about.jstor.org/terms
1798 White and Montes-R—Seed Germination of Common Bean
T able 3. Estimated values of base and optimal temperatures fo
germination, and maximum rate of germination for 16 genoty
evaluated at 11 temperatures
Values in parenthesis are asymptotic standard errors.

Identification Race" Temperature (°C) Maximum rate

of germination
Base Optimal (•/.d"1)
("/od"1)

7-3 (0-8) 27-4(0-6) A 36

70-0 (1-2) NG
8-0 (0-8) 26-2 (0-9) A 195
80-3 (3-5) NG
NG 7-2 (1-4) 27-2(1-1) A 197
69-9 (3-1)
NG 7-2 (0-9) 26-8(1-1) BAT 1393
88-5 (3-8)
ICA Linea 24 NG 81 (0-9) 26-4 (0-9) 108-2 (4-3)
WAF 10 NG 5-8 (2-3) 26-7(1-9) 64-7 (3-6)
A 22 MA 8-8 (0-4) 25-7(0-6) 84-9 (2-1)
BAT 881 MA 7-9 (0-8) 28-8 (0-7) 197-3 (3-2)
15 20 25 Carioca MA 8-9 (10) 25-5(1-1) 89-0 (4-0)
Jamapa
Temperature (°C) MA 8-4 (0-9) 28-0(1-0) 83-7 (2-4)
Magdalena 3 MA 8-7 (1-5) 27-5(1-5) 79-0 (5-2)
NEP Bayo 22 on
Fig. 3. Effect of temperature MA rate 28-5(1-1)
8-4 (0-7) of 100-6 (5-5)
germination
and A), and Sonora 32 (• Porrillo
and Sintetico
Sintético
A).MA 81 (10) circles
The 27-8 (0-9) 86-5 (O
(2-2) and •
(A and A) indicate observedSan Cristobal
data 83 MA from 26-8(1-0)
7-0 (0-5) separate 79-4 (3-3)replica
XAN 90 MA 8-5 (0-6) 29-3 (0-7) 80-9 (2-2)
Pinto UI 114 DG 8-5 (1-2) 27-4(1-2) 87-9 (4-2)
Table 2. Estimated values of base and optimal tem
germination, and 'DG = Durango,
maximum rate MA = Mesoamerican, and NG = Nueva Granada.
of germination for
evaluated in conventional incubators

Values in parenthesis are asymptotic standard errors.

30
Identification Temperature (°C) Maximum rate
of germination
Base Optimal (% d-1)
(yod"1) 9 29 A A
A A
Q) A
BAT 477 7-4 (0-5) 28-9 (0-5) 87-0 (1-2)
BAT 1297 6-8 (0-8) 28-7 (0-7) 76-3 (1-3) 1 28 A
A
0)
Diacol Calima 7-6 (0-6) 27-0 (0-4) 64-1 (0-9) Q.

Sonora 32 E
8-3 (0-6) 34-5 (1-4) 130-1 (5-3)
& 27
Am
CO
/A#
E
S. C. 83 »
O 26
conclusions of Scully and Waines (1987), where five A
A
Carioca
accessions of tepary bean showed more rapid germination 25
at 32 °C than five common bean genotypes.
6 7 8 9 10
In the comparison of 16 genotypes, Tb varied from
Base temperature
5-8 °C to 8-9 °C while T0 ranged from 25-5 °C to 29-3 °C
(Table 3). Comparisons based on standard errors again Fig. 4. Comparison of ba
suggested no differences in Tb, but that T0 differed among tion for 19 common be
genotypes. Granada (•), Mesoameri
pooled standard errors of
No difference between the Nueva Granada and
Mesoamerican races were found for Tb or T0 reflecting
ation
the large variation within races (Fig. 4). Nonetheless, for in temperatur
a mean germination temperature, calculated as the T0 mean appear to be insen
of Tb and Ta, the difference between the two races 1987; was Ellis and Butche
(Ellis and Butcher, 1
significant (PcO-Ol). The responses of the Mesoamerican
cultivars Carioca and San Cristobal were of particular BAT 881 stood out f
interest since their values of Ta were close to of those of germination (197
genotypes from the Nueva Granada race. Although small-seeded
these genoty
two cultivars are classified in the Mesoamerican race,
germination than la
their seed coloration patterns and other traitsno suggestassociation betw
they originated through inter-racial crosses with of thegermination. Lar
Nueva Granada race. Since all seed lots were obtained highest rate of the
from a single location and generally showed highseed final lot affects rate
percentage germinations, it seems unlikely that seed maximum
lot rate of
effects would have contributed significantly to the vari
attributed to the app

This content downloaded from

132.174.251.190 on Thu, 07 Sep 2023 02:55:56 +00:00
All use subject to https://about.jstor.org/terms
 White and Montes-R—Seed Germination of Common Bean 1799

lots. The results thus support the possibility of improving within the Mesoamerican and Andean gen
rate of germination regardless of seed size. although there are still differences between
The selection of genotypes for this study was con The single tepary bean accession evaluated,
strained by the desire to use seed of uniform provenance had the highest values of T0. The methodol
and which would be free of seed-borne pathogens. This appeared satisfactory for characterizing lar
precluded inclusion of genotypes adapted to highland germplasm, assuming that seed of good qua
sites which do not grow well at the Palmira station of able, although there is a need to clarify the poss
CIAT. Thus the genetic variation described in this study of growing environment or seed lot on the
probably does not represent the full range available in temperature response of seed germination.
common bean.
Although Tb and Ta were estimated using methods
somewhat different from those of Covell et al. (1986) for
LITERATURE CITED
chickpea, lentil, soyabean, and cowpea, the values of Tb
Brain P, Butler R. 1988. Cumulative count data. Genstat
and T0 for common bean appear consistent with expecta
Newsletter 22, 38—45.
tions for comparisons with the other species. The mean
Covell S, Ellis RH, Roberts EH, Summerfield RJ. 1986. The
value of Tb (7-8 °C) indicated that bean requires warmer
influence of temperature on seed germination rate in grain
temperatures than chickpea (0-0 °C), lentil (2-5 °C) and
legumes. I. A comparison of chickpea, lentil, soybean, and
soybean (4-0 °C), but that it tolerates cooler temperatures cowpea at constant temperatures. Journal of Experimental
Botany 37, 705-15.
than cowpea (8-5 °C). Values of T0 of common bean
Dickson MH. 1971. Breeding beans, Phaseolus vulgaris L. for
(25-5 °C to 29-3 °C) were higher than lentil (24-0 °C to improved germination under unfavourable low temperature
24-4 °C), but lower than the other species (minimum of conditions. Crop Science 11, 848-50.
31-8 °C for chickpea). This comparison thus suggests that Ellis RH, Butcher PD. 1988. The effect of priming and
common bean has a narrower range of adaptation for 'natural' differences in quality amongst onion seed lots on
the response of germination to temperature and the identifica
seed germination temperature than the other legumes.
tion of the characteristics under genotypic control. Journal of
The simulation model BEANGRO assumes a value of
Experimental Botany 39, 935-50.
Tb of 5 °C for all developmental processes and values Ellis of
RH, Covell S, Roberts EH, Summerfield RJ. 1986. The
30 °C and 18 °C (according to ecotype) for Ta of vegetative
influence of temperature on seed germination rate in grain
development (Hoogenboom et al., 1991). These valueslegumes. II. Intraspecific variation in chickpea (Cicer
indicate a large discrepancy with our values of Tb arietinum
and L.) at constant temperatures. Journal of
Experimental Botany 37, 1503-15.
T0. Thus, either seed germination shows a temperature
Ellis RH, Hong TD, Roberts EH. 1987. Comparison of
response that is distinct from other phases of vegetativecumulative germination and rate of germination of dormant
development, or the inconsistency is due to the differentand aged barley seed lots at different constant temperatures.
methods used to estimate the two sets of coefficients. Seed Science and Technology 15, 717-27.
The genotype showing the strongest effect of temper Ellis RH, Roberts EH. 1981. The quantification of ageing and
ature on maximum germination was WAF 10, where only survival in orthodox seed. Seed Science and Technology
9, 373^109.
23% germination was obtained in one replicate ofGarcia-Huidobro
the J, Monteith JL, Squire GR. 1982. Time,
34-5 °C treatment. However, the highest germination in
temperature and germination of pearl millet (Pennisetum
the replicate was only 77% (at 12-0 °C) suggesting typhoides
that S. & H.). I. Constant temperature. Seed Science and
Technology 33, 288-96.
poor seed quality may exacerbate effects of temperature
Genstat
at extreme temperatures. Further evidence for this was V Committee. 1987. Genstat 5 reference manual. UK,
Oxford: Clarendon Press.
provided by NEP Bayo 22, where germination at 34-5 °C
Hardwick RC, Andrews DJ. 1980. A method of measuring
was 86% and 69% in the two replicates, and maximum
differences between bean varieties in tolerance to sub-optimal
among all temperatures were 98% and 97%, for temperatures.
the Annals of Applied Biology 95, 235—47.
Hoogenboom
respective replicates. For all other genotypes, 100% ger G, McSmith DS. 1992. Temperature sensitivity
of common bean during germination. Annual Report of the
mination was always achieved for at least one temperature
in each replicate. The data of Ellis et al. (1986) Bean
for Improvement Cooperative 35, 72-3.
Hoogenboom G, White JW, Jones JW, Boote KJ. 1991.
chickpea genotypes L 550 and K 850 suggest a similar
Beangro VI 01 dry bean crop growth simulation model user's
effect of seed quality on maximum germination, and the
guide. Florida Agricultural Experiment Station Journal No.
results agree with the basic principles of seed ageingN-00379.
and
survival as described by Ellis and Roberts (1981). Hunter EA, Glasbey CA, Naylor REL. 1984. The analysis
of data from germination tests. Journal of Agricultural Science
CONCLUSION 102, 207-13.
Laing DR, Jones PG, Davis JHS. 1984. Common bean
Our study confirms previous reports of cultivar differences
(Phaseolus vulgaris L.). In: Goldsworthy PR, Fisher NM, eds.
in the response of germination to temperature, and further
The physiology of tropical field crops. New York: Wiley &
Sons, 305-51. exists
suggests that considerable variation in response

This content downloaded from

132.174.251.190 on Thu, 07 Sep 2023 02:55:56 +00:00
All use subject to https://about.jstor.org/terms
1800 White and Montes-R—Seed Germination of Common Bean
Ritchie JT, Ne Smith DS. 1991. Temperature and crop develop Scully B, Waines JG. 1987. Germination and emergence
ment. In: Hanks J, Ritchie JT, eds. Modelling plant and soil response of common and tepary beans to controlled temper
systems. ASA, CSSA, SSSA, Madison, Wisconsin, 5-29. ature. Agronomy Journal 79, 287-91.
SAS Institute Inc. 1985. SAS user's guide: Statistics, version 5 Singh SP. 1989. Patterns of variation in cultivated common
edition. Cary, N.C: SAS Institute Inc. bean (Phaseolus vulgaris, Fabaceae). Economic Botany 43,
Schoonhoven Av, Pastor-Corrales MA. 1987. Standard 39-57.
system for the evaluation of bean germplasm. CIAT, Cali, Singh SP, Gepts PA, Debouck DG. 1991. Races of common
Colombia. bean (Phaseolus vulgaris, Fabaceae). Economic Botany 45,
Scott S J, Jones RA, Williams WA. 1984. Review of data 379-96.
analysis methods for seed germination. Crop Science 24, Suggs CW, Mohapatra SC, Peel HB. 1987. Solid state cooled
1192-9. thermogradient incubator. Transactions ASAE 30, 1149-52.

This content downloaded from

132.174.251.190 on Thu, 07 Sep 2023 02:55:56 +00:00
All use subject to https://about.jstor.org/terms