FRANK C. SORENSEN AND ROBERT K. CAMPBELL
Forestry Sciences Laboratory, Pacific Northwest Fori's! and Rant;e Experiment Station, Forest Sen•ice, United Stales Department
1!/'Agricu/ture, Con•al/is, OR, U.S.A.
9733 I
Received February 20. 1978
SORENSEN. F. C., and R. K. CAMPBELL. 1978. Comparative roles of soil and air temperatures in
the timing of spring bud flush in seedling Douglas-fir. Can. J. Bot. 56: 2307-2308.
Outdoor cold frame tests in which air and soil temperatures were modified showed both to be
involved in the timing of spring bud flush. The results indicated that warming the soil advanced
the date of growth initiation and warming the air increased the rate of bud development once
initiated. The latter had a 5.7:0.45 greater effect per degree celsius change in temperature.
SORENSEN. F. C., et R. K. CAMPBELL. 1978. Comparative roles of soil and air temperatures in the
timing of spring bud flush in seedling Douglas-fir. Can. J. Bot. 56: 2307-2308.
Au cours d'experiences effectuees dans des couches froides, il s'est avere que des modifica­
tions des temperatures de !'air et du sol affectent toutes les deux Ia precocite de l'ouverture des
bourgeons. Les resultats font voir qu'un rechauffement du sol avance Ia date du debut de Ia
croissance et qu'un rechauffement de Ia temperature de l'air augmente Ia vitesse de developpe·
ment des bourgeons.une fois debourres. Cette derniere a un effet plus grand de 5.7:0.45 par degre
celsius de changement de temperature.
[Traduit par lejournal]
Introduction
cables placed 15 em below the soil surface and set to give an
species such as Douglas-fir (Pseudotsuga
menziesii (Mirb.) Franco), which show a chilling
requirement (Wommack 1964). bud-burst phenol­
ogy is considered to be determined lai·gely by rising
temperatures in the spring (Wareing 19S6). How­
ever, there is a question as to the relative impor­
tance of air and soil temperatures. Lavender eta!.
(1973) suggested that soil temperature played a
major role, with cold soils (SoC) delaying bud flush.
Timmis and Worrall (1974) observed no differences
in flushing date between cold- and warm-root
plants.
In 1970 and 197S, two experiments were estab­
lished in raised nursery beds in Corvallis, OR, in
which air and soil temperatures were artificially
raised. Both experiments showed a combined
influence of air and soil temperatures.
In
Materials and Methods
Four temperature treatments. normal and heated air and (or)
soil, were applied factorially and at random in a 2
x
2 arrange­
ment in both experiments. All treatments were replicated twice.
Plot size was 1.2
x
1.5 m. In experiment I (El), plots included 5
seedlings per provenance from 40 provenances; in experiment II
(Ell). plots included 12 seedlings
per provenance from 4 prove­
nances. Provenances sampled Douglas-fir stands between 43
and 48°
N latitude in western Oregon and Washington.
Warming of the air was achieved by placing a plastic tent over
the plots. To permit air circulation into the tents. bottom edges
of the plastic were supported 5 em above the soil surface and
ends of the tent frames were extended 10 em beyond the
framework of the raised beds. Soil was warmed with heating
average temperature of 20°C, IDem below the soil Stll·face. The
temperature treatments were started when the seeds were sown.
Temperatures were monitored in Ell during the 6 to 8 weeks
preceding and during bud flush. Air temperature was measured
2 em above and soil temperature 10 em below the soil surface.
Presence or absence of a flushed terminal bud was recorded
on every seedling twice weekly (EI) and every other day
(Ell). A
bud was considered flushed when green needles could first be
seen between the bud scales. Mean date of flushing was calcu­
lated for each plot. and plot means were subjected to analysis of
variance.
Results and Discussion
Air temperature in the covered plots averaged
2.1oc warmer than in the uncovered plots over the 4
to 6 weeks prior to bud flush and 2.9°C warmer
during the actual flushing period. Temperature dif­
ferences between covered and uncovered plots var­
ied with time of day and were greatest about 6 pm
(about SoC) and least about 6 am (about O.SOC). Soil
heat did not measurably affect air temperature.
Unheated soil warmed linearly with time from
mid-February and remained constant with time in
heated plots. Plastic tents raised soil temperature
by about 4 and l°C in soil-heated and unheated
plots, respectively. During the 30 days prior to
average date of bud flush for all plots, average soil
temperatures of the soil-heated and unheated plots
differed by 11.7°C.
Average date of terminal bud flush in EI was
advanced by 7.7 and 3.S days by increased air and
soil temperatures, respectively. The comparable
CAN.
230B J.
BOT VOL. 56.
figures for El l were 14.3 and 5.3 days. In Ell, in
which temperatures were monitored. flushing was
advanced 5.7 days per l°C increase in average daily
air temperature and 0.45 days per I oc increase in
average soil temperature.
The effects of air temperature (F = 46.35, p <
1978
usually of more importance in determining flushing
date than is the time at which measurable develop­
ment may have started.
We believe that several preconditioning factors
which influence developmental rates have often
been neglected or inadequately characterized when
O.Oi,df, L7 inEI;F= 1!4.08,p<O.O l,df.L4 in
the influence of individual environmental treat­
Ell) and soil temperature (F= 9.68.p<0.05 in EI;
ments on bud-flush phenology has been studied.
F = 15.65. p < 0.05 in Ell) were both significant.
These factors. in addition to chilling duration and
Interaction between air and soil temperatures was
not significant in either experiment.
photoperiod (Flint 1974; Campbell and Sugano
1975), include chilling temperature and length of
Time of bud flush depends on the initiation of a
time between bud set and initiation of chilling
growth process and the rate at which growth pro­
(Dormling eta!. 1968; Sugano 1971). We also be­
ceeds once initiated. Initiation has received the
lieve that the particular importance of these factors
major attention in previous physiological studies of
is their influence on developmental rate or response
bud flush, but we believe our observations on the
to heat-unit accumulation at low temperatures
differential effect of air and soil temperatures indi­
(Vegis 1963). Much of the development toward bud
cate that rate of development,rather than the time
flush normally occurs in the moderately low tem­
peratures characteristic of early spring (5-10°C). It
of initiation. has greater influence on bud-burst tim­
mg.
Lavender et a!. (1973) advanced bud flush by
about 2 weeks by increasing soil temperature from
5 to 20°C or 0.9 days/l°C. Their temperature treat­
ments were started about 6 weeks before bud flush
and were applied to seedlings which had been main­
tained out of doors all winter until late February.
Day lengths were short (9 h) and air temperature
was a constant 20°C. In our experiments, at the
time of bud flushing, day lengths were intermediate
(about 13 h), average air temperature was about
i0°C. contrasting soil temperatures about 9 and
2!°C, and the plants had been outdoors until the
time of bud flush or until about mid-April.
While results of both tests show that soil temper­
ature plays a role in bud-burst phenology, our re­
sults indicate that the role is negligible unless other
external factors are favorable. In winter and early
spring, response to temperature is limited by in­
adequate chilling and short photoperiods (Flint
1974; Campbell
and Sugano 1975).
Later,
the
influence of temperature increases as other factors
become less limiting (Brouwer 1964; Brouwer and
Kleinendorst 1967). Consequently, even though
soil temperature in soil-heated plots was 21oc
throughout the winter, bud flush was delayed until
spring, in fact, until just shortly before bud flush in
unheated plots.
was at these temperatures that we found a small
difference in air temperature to have a large effect
on bud-burst timing.
BROUWER, R. 1964. Responses of bean plants to root tempera­
tures. I. Root temperatures and growth in the vegetative
stage.
g
Inst. Bioi. Scheikd. Onderz. Landbouwgewassen
Wa eningen Jaarversl. 12: 11-22.
BROUWER, R., and A. KLEINENDORST. 1967. Responses of bean
plants to root temperatures. Ill. Interactions with hormone
treatments. Inst. Bioi. Scheikd. Onderz. Landbouwgewassen
Wageningen Jaarversl. 15: l l -28.
CAMPBELL, R. K., and A. I. SUGANO. 1975. Phenology of bud
burst in Douglas-fir related to provenance, photoperiod. chil­
ling and flushing temperature.
Bot. Gaz. (Chicago). 136:
290-29B.
DoRM LING.!.. A. Gus·I AFSSON. and D. VON WETTSTEIN. l96!l.
The experimental control of the life cycle in Picea abies (L.)
Karst. I. Some basic experiments on the vegetative cycle.
Silvae Genet. 17: 44-64.
FuN·I. H. L. 1974. Phenology and genecology of woody plants.
In Phenology and seasonality modeling. Edited by Helmut
Lieth. Springer-Verlag. New York. pp. 83-97.
LAVENDER, D.P .. G. B. SWEET, J. B. ZAERR, and R. K. HER­
MANN. 1973. Spring shoot growth in Douglas-fir may be in­
itiated by gibberellins exported from the roots. Science. 182:
838-839.
SUGANO, A. I. 1971. The effects of low temperatures on dor­
mancy release in Douglas-fir (Pseudotsuga menziesii (Mirb.)
Franco) from western Oregon. Washington. and California.
M.S. thesis, Oregon State University. Corvallis. OR.
TIMMIS, T., and J. WoRRALL. 1974. Translocation of deharden­
ing and bud-break promoters in climatically ·split' Douglas­
fir. Can. J. For. Res. 4:229-237.
Under natural and most nursery conditions, the
VEGIS, A. 1963. Climatic control of germination. bud break, and
effect of early initiation of bud activity is probably
vidually or in combination to slow the rate of de­
dormancy. In Environmental control of plant growth. Edited
bi' L. T. Evans. AcademicPress. New York. pp. 265-285.
w,;REING. P. F. 1956. Photoperiodism in woody plants. Annu.
Rev. PlantPhysiol. 7: 191-214.
WoMMACK, D. E. 1964. Temperature effects on the growth of
velopment or response to heat-unit accumulation.
Douglas-fir.Ph.D. thesis, Oregon State University. Corvallis,
Therefore, it appears that developmental rate is
OR.
inconsequential because lack of chilling, short
photoperiods, and low air temperatures act indi­