Plant Physiol. (1986) 81, 1075-1079 0032-0889/86/81/1075/05/$01.00/0

Carbohydrate Level and Growth of Tomato Plants II. THE EFFECT OF IRRADIANCE AND TEMPERATURE Received for publication February 4, 1986 and in revised form March 17, 1986

MARTIN P. N. GENT Department of Forestry and Horticuhture, The Connecticut Agricultural Experiment Station, New Haven, Connecticut 06504 ABSTRACT The growth response of tomato (Lycopersicon esculentum L.) to temperature and irradiance may be related to carbohydrate concentration. Plants in the exponential phase of vegetative growth were grown under temperatures ranging from 9 to 36°C and under low or high irradiances of approximately 110 or 370 microeinsteins per square meter per second photosynthetically active radiation for a 12 hour photoperiod. The relative growth rate, leaf area ratio, net assimilation rate and whole plant carbohydrate levels were measured. At high irradiance, relative growth rate was 43% faster and total nonstructural carbohydrate concentration was 41% greater than at low irradiance. The change in carbohydrate with irradiance could explain the growth response. Plant growth was fastest at 25°C and decreased parabolically at lower and higher temperatures with a half-maximal rate at 13 and 36°C. Total nonstructural carbohydrate decreased between 13 and 23'C and remained constant at higher temperatures. Soluble sugar concentrations varied little with temperature above 13°C except for sucrose, whose level rose above 30°C. The change in carbohydrate with temperature could not explain the growth response. Above 23°C tomato plants appeared to regulate growth rate to maintain a relatively constant nonstructural carbohydrate concentration.

bohydrate concentration, TNC, varies according to the rates of net CO2 exchange and synthesis of structural material (12).

dt

=

0.68 -(L.4-Rm-Rg RGR- 0.92 * RGR

(2)

Net CO2 exchange (the quantity in parentheses in equation 2) increases according to the rate of photosynthesis (P, g CO2* m2 h-'). Photosynthesis is assumed to saturate at high irradiance, CO2 and temperature with half maximum values at 190 gE. m-2 s PAR, 300 l L-' CO2 and 12°C (6). Photosynthesis per unit leaf area is normalized per unit dry weight of plant by LAR (g- m2). Net CO2 exchange decreases according to the rates of maintenance respiration (Rm, g CO2 -g ' h') and growth respiration (related to RGR by Rg, 0.68 g CO2 g-' structural material synthesis), both of which increase exponentially with temperature. TNC also decreases by transformation of nonstructural carbohydrate to structural material. The interaction between equations (1) and (2) is the heart of the model. The model predicts that at low temperature the rate of metabolism limits growth while at high temperature the rate of photosynthesis and the TNC concentration limit growth. As temperature increases from 12 to 23°C the model predicts photosynthesis increases to a plateau and the rate of metabolism doubles. The rate of metabolism no longer limits growth at 23°C and TNC decreases between 12 and 23°C. As temperature increases above 23°C photosynthesis remains constant and the rate of metabolism continues to increase, leading to very low TNC. The model predicts that more TNC is used for maintenance than growth at low TNC so RGR would decrease at temperatures above 23°C. This type of model was tested for plants grown under different temperatures, diurnally fluctuating and constant. It predicted photosynthesis and respiration in one experiment (11), and growth and diurnal variation in TNC in another (7). However, neither TNC nor RGR varied greatly between constant and fluctuating temperatures. A more stringent test of the model would be an examination of growth and carbohydrate concentration over a wide range of temperatures. This paper describes relative growth rate, net assimilation rate, and leaf area ratio measurements of tomato plants acclimated to a wide range of temperature and two levels of irradiance. The concentrations of TNC and various sugars that make up TNC were examined in plants grown under all conditions. -

The combined effects of irradiance and CO2 level on the relative growth rate of tomato plants have been explained by a whole plant model relating net CO2 assimilation to production of structural dry matter (14). A central assumption of the model is that a linear relation exists between total nonstructural carbohydrate concentration and RGR' (13 and references therein). Limited experimental evidence supports this hypothesis. Starch concentration of tomato leaves at the end of the photoperiod was highly correlated to RGR, although the sucrose concentration was not (8). In addition, an approximately linear relation was observed between the leaf extension rate and the carbohydrate concentration of cereals (12). The model has been extended to predict the temperature dependence of growth (6) by assuming an exponential increase with temperature in metabolism and growth (10, 11), doubling every 10°C. The exponential temperature dependence is multiplied by the linear carbohydrate concentration dependence to give RGR: RGR = K- TNC exp(0.0693 [ T - 25°C]) MATERIALS AND METHODS (1) where T is temperature and Kg (0.065 gg-'- TNC-'- h-') is the Plant Material and Growth Conditions. Hybrid tomato seed coefficient of growth metabolism. The total nonstructural car- (Lycopersicon esculentum L.) cv Sonato (Deruiter Zonen, Bleiswijk, Holland) was germinated in fine vermiculite at 23C and ' Abbreviations: RGR, relative growth rate; LAR, leafarea ratio; NAR, grown under an irradiance of 220 gE- m-2. s-' PAR and a 12 h net assimilation rate; TNC, total nonstructural carbohydrate. photoperiod. When the second true leaf was 1 cm long, uniform -

1075

1076

Plant Physiol. Vol. 81, 1986

GENT

seedlings were transplanted up to the cotyledons in fine vermiculite in 10 cm diameter pots and placed on benches in the center of the growth room. The growth room had a 12 h photoperiod with irradiance provided by an equal mix of standard and wide spectrum GroLux fluorescent bulbs spaced at 6 cm intervals across the ceiling. The upper bench, 60 cm below the lights, provided the high irradiance condition. The lower bench, 100 cm directly beneath the upper bench, provided the low irradiance condition. Irradiance measured at nine points on each bench using a quantum sensor (LI-COR model 190S) varied less than 10% across each bench. A fan circulated air turbulently within the growth chamber to equalize air temperature on both benches. Temperature, continuously recorded by thermograph on the lower bench, fluctuated 2°C with a cycle of 10 min during the day and 30 min during the night. CO2 concentration was 350 to 400 gl L' and RH was 20 to 30%. At dawn, pots were watered to excess with half-strength modified Hoagland solution warmed to the growth temperature. If the experimental temperature was more than 5°C different than the germination temperature, temperature was adjusted by 5°C/d at dawn until the desired condition was reached. The plants were acclimated for 3 d or until they had more than doubled in size before growth measurements commenced. Four harvests were made in each experiment. The first harvest was made at dawn. Growth continued for 3 or more d until plants had again more than doubled in size. Three consecutive harvests were then made at dawn, at dusk, and the following dawn. At each harvest every fourth row of plants was taken for a total of 16 plants from each bench. Pots were floated in water to saturate the vermiculite and roots were pulled free and washed. Plants were frozen and freeze dried. Roots, stems, and leaves of the dried plants were weighed, then these parts were combined and ground finely for carbohydrate analysis. At the final harvest, leaves were removed and sandwiched between cellophane sheets to measure total leaf area. In one experiment at 23°C, eight plants were harvested every 2 or 3 d over an 18 d period and leaf area was determined at each harvest. Growth Analysis. Relative growth rates were determined from the daily increase in natural log of whole plant dry weight. The initial, intermediate and final harvests at dawn were used for the determination. Error in determining RGR was 0.0 10 gg-' .d-' estimated from three growth experiments at 23°C. LAR was determined from whole plant dry weight at the final harvest divided by the leaf area. NAR was the product of RGR and LAR at the final harvest. Carbohydrate Analysis. The carbohydrate concentrations were analyzed in plants from the two intermediate harvests at dawn and dusk. Subsamples of 100 mg of whole plant tissue were extracted repeatedly with chloroform:methanol:water 5:12:3 by volume and the chloroform phase separated and discarded. The aqueous phase was dried at 5O°C under a stream of air and dissolved in 1.0 ml of water. Soluble sugars were analyzed by HPLC (7). Starch in the insoluble residue from the extraction was digested with a-amylase and analyzed as glucose equivalents using a colorimetric procedure (7). Total soluble sugars were also assayed by the colorimetric procedure. Carbohydrate analyses were repeated three times. Standard errors of carbohydrate analyses were 0.6, 0.3, 0.5, and 9.0 mgg-' for sucrose, glucose, fructose, and starch, respectively. RESULTS AND DISCUSSION Growth Analysis. Relative growth rate and weight ratios of plant parts were constant over the period of measurement. In an initial study, growth was exponential over 18 d at 23°C as plant weight increased from 0.002 to 0.5 g. In the first 8 d of this study, corresponding to the growth interval in later experiments, the

leaf area relative growth rate (cm2 cm2 d- ') was the same under high and low irradiance and equal to RGR under high irradiance. After 8 d of acclimation, the leaf area relative growth rate under low irradiance dropped to a value equal to RGR under low irradiance. Values for RGR were 0.333 and 0.238 gg-' d-', under high and low irradiance respectively, and remained constant for 18 d. These values of RGR were similar to values observed elsewhere for tomato plants grown at 21°C under similar irradiance, daylength, and CO2 level (8). Values of NAR of 13.8 and 4.8 g.m2 -d-' were substantially less than observed previously under similar conditions (8). This was due to a smaller LAR and leaf to whole plant weight ratio in the present study. Averaged over all temperatures, plants on the upper bench grew under an irradiance of 373 uE m 2.s-' while those on the lower bench received 111 1 E . m2-s'. Irradiance varied up to 15% between experiments, due to the temperature sensitivity and aging of the lights, but the experimental design ensured the ratio of irradiances on upper and lower benches remained the same. Average air temperature was the same on both benches. Relative growth rates for tomato plants grown under high irradiance was 43% greater than under low irradiance, averaged over all temperatures (Fig. 1). Most of the observed variation in RGR could be accounted for by multiple linear regression including terms in irradiance (I, gE * m2 s-'), temperature (T, C), and temperature squared. This accounted for 90% of the variance observed, i.e. R2 = 0.90. Predictions of the regression formula: RGR = -0.2949 + 0.000329.I + 0.04001. T - 0.000797. T2 (3) are also plotted (Fig. 1). Deviations from a smooth parabolic response of RGR to temperature were due to variation in irradiance between experiments. The regression coefficients related to temperature predicted a maximum RGR at 25°C with a symmetrical decrease at lower or higher temperatures to half maximum at 13 and 36C. The prediction was not improved by adding other terms in irradiance or temperature. There was no significant interaction between effects of irradiance and temperature. The lack of interaction was observed previously in a study of younger tomato seedlings (9). Plants grown under high irradiance had leaf area ratios 92% greater than under low irradiance, averaged over all temperatures (Fig. 2). Only at 1 3°C was LAR significantly greater than at other temperatures. Correlation between LAR and RGR was not sig-

A

4

0

od

*

0.3

cm 4; =

0.2

to

';

1

0.0OL 10

20

30

Temperature,'C FIG. 1. The effect of growth temperature and irradiance on the relative growth rate of tomato plants. (0), High irradiance; (0), low irradiance. SE was 0.010 g-g-'-d-' for three replicate experiments at 23°C. Continuous curves are predictions from the regression equation (3).

CARBOHYDRATE LEVEL AND GROWTH

c,j 60

6E od

,-40 40

-i

10

20

30

Temperature,.C FIG. 2. The effect of growth temperature and irradiance on leaf area ratio of tomato plants. (0), High irradiance; (0), low irradiance. ;S() C-v

I 0

ci

15

1

0

6 0

0

* 0~~~~~~~

c

S

Z

0~~~~

o

If

0

0

o

0o

2

tO

20

0

30

Temperature,'C FIG. 3. The effect of growth temperature and irradiance on the net assimilation rate of tomato plants. (0), High irradiance; (0), low irradiance.

1077

was some overlap in TNC concentration of plants grown at high and low irradiance; TNC of plants harvested at dusk under low irradiance was generally greater than TNC at dawn under high irradiance (Fig. 4). The diurnal variation in TNC, the difference between TNC in plants harvested at dawn and dusk, was always greater under high irradiance than under low irradiance. Starch accounted for 80 to 90% of the TNC at temperatures of 13°C and above. This carbohydrate pool was never totally depleted at any growth temperature. The greatest starch concentrations were observed at 13 and 18°C; 310 and 250 mgg-' under high irradiance and 225 and 195 mg-g-' under low irradiance, for plants harvested at dusk and dawn, respectively. The starch concentration was lower at temperatures above 20°C and decreased only slightly with increasing temperature. In the range of 21 to 36°C starch concentrations averaged 200 and 127 mg-g-' under high irradiance and 140 and 83 mg-g-' under low irradiance, in plants harvested at dusk and dawn, respectively. Soluble sugars accounted for an average of 12.5% of the TNC at temperatures of 1 3C and above. Glucose and fructose concentrations changed little with temperature in the range of 13 to 36°C except for a slight reduction above 30°C that was more marked in plants grown under low than high irradiance. Averaged over all samples in this temperature range, there was significant diurnal variation in soluble sugar concentrations under high irradiance but not under low irradiance. Glucose concentrations averaged 5.2 and 4.3 mgg-' under high irradiance and 4.6 and 4.4 mg-g-' under low irradiance, in plants harvested at dusk and dawn, respectively. Fructose concentrations averaged 11.2 and 9.6 mg-g-' under high irradiance and 8.4 and 7.9 mgg-' under low irradiance, in plants harvested at dusk and dawn, respectively. Plants grown at 9°C deviated from this pattern; fully half of TNC was in the form of soluble sugars. Glucose and fructose concentrations in plants grown at 9°C were about 80 mg-g-' and 120 mg.g', respectively, with little variation between dawn and dusk. Sucrose concentration responded differently to irradiance and temperature than did the other soluble sugars (Fig. 5). Sucrose concentration decreased slightly with increasing temperature for plants grown in the range of 13 to 28°C. In this temperature range the sucrose concentration was 12.2 and 6.8 mg-g-' under high irradiance and 7.5 and 5.2 mg-g-' under low irradiance, in plants harvested at dusk and dawn, respectively. At 9°C sucrose concentrations were higher, 17 to 20 mg - g', but not as high as

400 nificant. The leaf area measured at the final harvest for plants 0 not than that for irradiance was grown under high greater plants n I Y a grown under low irradiance; it was weight per unit leaf area and E weight per plant that changed with irradiance. The rate of leaf 4 - 300 w 0 area expansion was unaffected by irradiance, at least during 0 growth in the first 8 d or 4-fold increase of dry weight under * * different irradiance levels. Leaf to stem to root dry weight ratios .L. U (9 a were 0.668:0.210:0.112 for plants under high irradiance and 0 d 200 0.651:0.252:0.097 for those under low irradiance. These ratios 0 varied little with temperature. 00 Plants grown under high irradiance had a net assimilation rate 0 2.75-fold greater than under low irradiance, averaged over tem100 |' 0 peratures of 13C and above (Fig. 3). There was a marked I0 decrease in NAR at 9C. The variation between experiments in NAR, that was not accounted for by irradiance and temperature, was less than that for LAR. This could indicate compensation 20 30 l0 between LAR and photosynthesis per unit leaf area. Temperature,LC Carbohydrate Analysis. Plants grown under high irradiance had TNC concentration 40.6% greater than those under low FIG. 4. The effect of growth temperature and irradiance on the total irradiance, averaged over dawn and dusk measurements at all nonstructural carbohydrate concentration of tomato plants. (0), High temperatures. This difference was similar to that observed for irradiance harvested at dawn; (-), high irradiance harvested at dusk; (0), RGR. TNC decreased with increasing temperature to 20C then low irradiance harvested at dawn; (0), low irradiance harvested at dusk. remained approximately constant at higher temperatures. There SE was 10 mg-g-' for three replicate samples. I

-

o

.~~~~~~~

0

1078

GENT 20

La 0

0

*

10 *

o

0

i

co

0~~~~ 50

0

0 1

0

20 Temperature,*C

30

FIG. 5. The effect of growth temperature and irradiance on the suconcentration of tomato plants. (0), High irradiance harvested at dawn; (e), high irradiance harvested at dusk; (0), low irradiance harvested at dawn; (0), low irradiance harvested at dusk. SE was 0.6 mg-g-' for three replicate samples. crose

observed for the other soluble sugars. Unlike the other soluble above 28C sucrose concentration also rose, increasing about 2-fold by 36C. This increase was more marked in plants harvested at dusk than dawn. Relation between Carbohydrate Concentration and Growth Rate. The model describing the relation between TNC and RGR has two assumptions. The first is that, at a given temperature, RGR is proportional to TNC. This hypothesis was supported by the results. The percentage increase in RGR with irradiance was similar to the percentage increase in TNC, averaged over all temperatures. A study that varied RGR and TNC through variation of irradiance, daylength, and CO2 concentration also showed a proportional relationship (8). The proportionality between TNC and RGR observed here was the same as that observed for plants growing in a greenhouse under constant or fluctuating temperature with a mean of 1 7C (7). However, both RGR and TNC observed here were 2 times the level observed in greenhouse plants grown at a similar temperature. Early work also found a linear relation between sucrose concentration taken up by a leaf and growth of the stem apex in the dark (1). This effect was independent of growth temperature. The second assumption of the model is that, at a given TNC concentration, growth increases exponentially with temperature, doubling every 10°C. The results did not support this. RGR divided by TNC gave the temperature dependence of growth sugars,

Plant Physiol. Vol. 81, 1986

independent of TNC, and it was not exponential. The temperature dependence of RGR/TNC was a symmetric parabola, such as that describing the temperature dependence of RGR. The only difference was a skew toward slightly more rapid growth at temperatures between 25 and 35C. The regression equation that best fit RGR/TNC gave a temperature optimum of 28°C with a half maximum at 16 and 40°C, R' = 0.88. There was no

significant correlation between irradiance and RGR/TNC. There were other differences between the predictions of the model and the observed RGR and TNC at dusk and dawn. Observed and predicted values are given in Table I. Measured values of irradiance, CO2 concentration, day and night temperatures, and leaf area ratio were used in the simulation. The coefficient for growth metabolism was 0.065 g CO2g-' *TNC-' h-'. With these exceptions the equations of the model and the parameters used to predict RGR and TNC were as described earlier (6). The model predicted that RGR would rise between 13 and 23°C then remain constant to 33°C. It predicted TNC at dusk would remain constant between 13 and 23°C then fall to 33C. The observations contradicted these predictions. RGR decreased between 23 and 33°C and TNC decreased between 13 and 23C. At all temperatures the model predicted a greater diurnal variation in TNC than was observed. When it was not limited at low temperature by the rate of metabolism, the model predicted rapid growth throughout the diurnal cycle until TNC was almost depleted. A more accurate summary ofthe observations was that RGR was controlled by the plant in order to maintain a more constant TNC. TNC never fell to a level representative of carbohydrate starvation that might limit growth, such as observed for cucumber grown under low irradiance at 25°C (2). The model failed to predict the greater TNC and RGR in plants grown under high irradiance than in those grown under low irradiance. According to the model there was complete compensation between photosynthesis per unit leaf area and LAR so that photosynthesis per unit dry weight and therefore TNC and RGR were unchanged by irradiance. The observations showed a high degree of compensation that reduced the response of TNC to irradiance but compensation was not complete. Whereas irradiance and NAR were 3-fold greater for tomato plants grown at high than low irradiance, LAR and the specific leaf area compensated so that the increase in RGR and TNC was only 43% greater, or one-fifth of the primary response. Acclimation of photosynthesis to irradiance would tend to increase photosynthesis per unit leaf area for plants grown under high irradiance compared to those grown under low irradiance. In fact, deviations by the model from the observations could be explained by a greater rate of photosynthesis under high irradiance than predicted by the model. Those instances where growth conditions resulted in carbo-

Table I. Relative Growth Rate and Total Nonstructural Carbohydrate Concentration in Tomato Plants Observations were compared to predictions of the model. Total Nonstructural Carbohydrate Relative Growth Growth Conditions Rate Concentration Dawn Dusk Irradiance Temperature Observed Predicted Observed Predicted Observed Predicted M-2 g g_, vd-' °C s' g-g-' AE. 0.224 0.142 0.226 0.126 0.225 0.144 12.8 122 0.116 0.094 0.245 0.183 0.235 0.238 22.5 120 0.105 0.014 0.174 0.159 0.265 0.202 31.4 118 0.153 0.287 0.241 0.331 0.135 0.194 12.8 378 0.149 0.102 0.260 0.249 0.251 0.333 22.5 409 0.011 0.170 0.158 0.226 0.292 0.236 31.4 384

CARBOHYDRATE LEVEL AND GROWTH hydrate concentrations outside the usual range were at temperatures close to the limits of acclimation for tomato. Cold or heat acclimation was constant when tomato plants were grown in the range between 13 and 26°C (5). I saw no change in soluble sugar concentration in this temperature range. At lower temperatures a slight hardening to chilling injury was found (5), that may be related to the increase in soluble sugars I noted at 9°C. This indicated a failure to synthesize starch, since starch concentration fell between 13 and 9°C while soluble sugar concentrations rose. The starch synthase enzyme was found to be temperature sensitive, and showed a decrease in activity below 12°C in chilling sensitive plants (4). Above 30°C, tomato was partially hardened to heat injury (5). This may be related to the increase in sucrose concentration noted here. Above 30°C, an increase in sucrose was seen in tomato flower buds (3), and it was attributed to a reduction in activity of the enzyme invertase. In the whole plant there was no concomitant reduction in glucose and fructose, corresponding to the rise in sucrose above 30°C. This suggested that invertase did not limit the metabolism of carbohydrate at high temperature in the whole plant. In summary, during exponential growth of tomato plants in the vegetative phase, compensatory mechanisms were at work to keep the carbohydrate concentration as constant as possible over a wide temperature range. The growth rate, among other aspects, was modified to maintain carbohydrate concentration relatively constant throughout the temperature range of acclimation. LITERATURE CITED 1. BOHNING RH, WA KENDALL, AJ LINCK 1953 Effect of temperature and sucrose on growth and translocation in tomato. Am J Bot 40: 150-157

1079

2. CHALLA C 1976 Analysis of the diurnal course of growth, carbon dioxide exchange, and carbohydrate reserve content of cucumber. Agric Res Rep Wageningen 861 3. DINAR M, J RUDICH 1985 Effect of heat stress on assimilate metabolism in tomato flower buds. Ann Bot 56: 249-257 4. DOWNTON WJS, JS HAWKER 1975 Response of starch synthesis to temperature in chilling-sensitive plants. In R Marcelle, ed, Environmental and Biological Control of Photosynthesis. Dr W Junk, The Hague, pp 81-88 5. DROZDOV SN, AF TITov, VV TALANOVA, SP KRITENKO, EG SHERUDILO, TV AKIMOVA 1984 The effect of temperature on cold and heat resistance of growing plants. I. Chilling sensitive species. J Exp Bot 35: 1595-1602 6. GENT MPN, HZ ENOCH 1983 Temperature dependence of vegetative growth and dark respiration: a mathematical model. Plant Physiol 71: 562-567 7. GENT MPN 1984 Carbohydrate level and growth of tomato plants. I. The effect of carbon dioxide enrichment and diurnally fluctuating temperatures. Plant Physiol 76: 694-699 8. HURD RG, JHM THORNLEY 1974 An analysis of the growth of young tomato plants in water culture at different light integrals and CO2 concentrations. I. Physiological aspects. Ann Bot 38: 375-388 9. HUSSEY G 1965 Growth and development of the young tomato; III. The effect of day and night temperatures on vegetative growth. J Exp Bot 16: 373-385 10. MCCREE KJ 1974 Equations for the rate of dark respiration of white clover and grain sorghum as functions of dry weight, photosynthetic rate and temperature. Crop Sci 14: 509-514 1 1. MCCREE KJ, ME AMTHOR 1982 Effects of diurnal variation in temperature on the carbon balances of white clover plants. Crop Sci 22: 822-827 12. PENNING DE VRIES FWT, JM WITLAGE, D KREMER 1979 Rates of respiration and of increase in structural dry matter in young wheat, ryegrass and maize plants in relation to temperature, water stress and to their sugar content. Ann Bot 44: 595-609 13. THORNLEY JHM 1982 Interpretation of respiration coefficients. Ann Bot 49: 257-259 14. THORNLEY JHM, RG HURD 1974 An analysis of the growth of young tomato plants in water culture at different light integrals and CO2 concentrations. II. A mathematical model. Ann Bot 38: 389-400