American Journal of Botany 45: 626-631 (1958)
H. R. Highkin

1Received for publication March 28, 1958.
2Supported in part by a grant from the National Science Foundation (G-1385).

SPECIFIC ENVIRONMENTAL factors such as photoperiod and thermoperiod and their effect on reproduction, fruit set, etc. have been extensively studied (Went, 1948; Lang, 1952). Went (1948, 1953), in his work on thermoperiodicity, has pointed out that some plants require variations in temperature between day and night in order to obtain optimal growth and fruit set.

However, the relationship between a controlled environment, both constant and fluctuating, and the growth and development of higher plants has not been studied in great detail. This is particularly true with regard to those characteristics which appear to he quantitative in nature and which are most readily influenced by the environment. Moreover, the relationship between these environmental factors and the characteristics of the offspring has not been studied at all. This has been due primarily to the lack of adequate facilities. Since the construction of the Earhart Plant Research Laboratory at the California Institute of Technology such facilities have become available. As a result, it is primarily quantitative characters such as growth rate, yield, heat. and cold-resistance, leaf size and shape and general vigor, which have been and are being studied in this laboratory from a physiological as well as a hereditary point of view.

The present studies were designed to determine the effect of controlled environment on the growth and development of plants over a number of generations. Two important facts have emerged in regard to growth of peas: (1) A constant temperature is inhibitory for growth despite the fact that it may be the most nearly optimal constant temperature; furthermore, the inhibitory effect is cumulative from generation to generation reaching saturation with approximately the 5th generation. (2) Differences in the temperature during growth, whether constant or diurnally fluctuating, result in differences in the growth characteristics of the offspring of plants grown in these controlled conditions.

MATERIALS AND METHODS.The plants used in all these experiments were two varieties of peas, namely, 'Unica,' a commercial variety obtained from Holland, and 'L5' obtained from W. Lamprecht in Sweden. Both varieties have been inbred since the early 1930's and are very uniform. Because of the structure of its flower and because pollination takes place even before the flower bud opens, the pea is completely self-pollinated. It is therefore most likely that these two varieties, which have gone through an inbreeding and selection program extending for some 20-25 years, are relatively completely homozygous.

All the experiments were performed in the Earhart Plant Research Laboratory over the last four years. A description of this laboratory and of the experimental facilities available in it has been published by Went (1957).

The original seeds from which all the present seeds descended were planted in either constant temperature-artificial light chambers, or in controlled fluctuating temperatures in greenhouses with natural light supplemented by low intensity artificial light in order to make a 16-hr. photoperiod. This 16-hr. photoperiod was used in all conditions.

The three fluctuating temperatures used were: (1) constant day temperature of 17C.constant night of 11C.; (2) constant day temperature of 20C.constant night of 14C.; (3) constant day temperature of 23C.constant night of 17C.

Precautions were taken to avoid selection in favor of any given environment in order to see any mutation which might occur and trace it to its origin. Eighteen seeds of each variety were planted in each constant temperature chamber and greenhouse. In each generation, one seed from the first pod of each of the 18 plants was used in continuing the population so that in effect each strain represented a series of 18 pure lines.

Fig. 1. Effect of the number of generations at a constant temperature (10C.) on the growth of pea plants. For each variety, seeds from plants grown for 0, 1, 3, and 5 generations at constant 10C. were grown simultaneously to obtain curves. The data given represent daily increase in height (mm.) during linear phase of growth.

RESULTS.Inhibitory effects of a constant temperature.A line of plants was grown for a number of generations at constant temperatures (10C., 17C., and 20C., respectively), with artificial light of approximately 1000 ft.-c.

Growth rate measurements at each constant temperature revealed a steady decrease in vigor in successive generations. Fig. 1 shows the results for both varieties grown at 10C. and Fig. 3 compares the first and fifth generations at the same temperatures. Results from the sixth generation onward are not reported, since the vigor had decreased to such an extent that seeds from these plants were practically inviable. They were extremely susceptible to rot and those few seedlings which were obtained were abnormal, the growing points dying after three or four nodes had been formed.

These results are characteristic of those obtained at temperatures of 17C. and 20C. The experiments have been repeated at least three times at each temperature with essentially the same results. To obtain the curves shown in fig. 1, seeds from plants grown for 0, 1, 3, or 5 generations at 10C. were planted at the same time. Accompanying the decrease in vigor shown in fig. 1, there was a decrease in both the number and viability of the seeds. The seeds for the "controls," that is those plants growing at the constant temperature for the first time and used as a basis of comparison, were harvested from plants grown under controlled fluctuating temperatures in greenhouses.

Reversibility of the inhibitory effects of a constant temperature.The constant temperature inhibition of vigor is reversible by returning the line to fluctuating temperature conditions of the greenhouse. However, complete reversibility of constant temperature inhibition requires at least three generations' growth under the fluctuating temperature condition. To assay for the reversibility of the constant temperature inhibition of vigor, a line which had grown for 5 consecutive generations at constant temperature was grown for 1, 2, or 3 generations, respectively, under fluctuating temperature conditions and then returned for one more generation's growth at the constant temperature. The vigor of the line was measured relative to controls in both the fluctuating conditions and the subsequent constant temperature condition. It was found that when the vigor of these lines was measured under the fluctuating conditions, almost complete reversal appears to have occurred. However, when the line was returned to constant temperature conditions for assay, reversal was only partial and was proportional to the number of generations that the line had spent in the fluctuating condition. The results are depicted schematically in fig. 2, while data for one variety ('Unica') at one constant temperature are given in table 1. With regard to the data there presented, it should be noted that there are differences in the final height of the controls in the successive generations. This is due to uncontrollable seasonal variation in the natural conditions, of which light quality as well as quantity are known to vary within the greenhouses. The only controls possible at present in the greenhouses are those which control the ambient temperature and the humidity. In all these experiments extreme care was taken to ensure uniformity of environmental conditions between groups during any single experiment. Comparison can be made, therefore, only between the "treated" plants and their corresponding controls within any single experiment. This is true in these experiments for the constant temperature-artificial light conditions as well. Plants were randomized to ensure uniformity of conditions at any given constant temperature. At any given constant temperature the light conditions were only approximately the same between experiments, i.e., between successive generations. Here, again, comparison can be made only between "treated" plants and their corresponding controls within any single experiment.

TABLE 1. Reversibility of constant temperature growth inhibition for
a line grown for 5 generations at a constant temperature of 10C.

a) Assay in fluctuating temperature of greenhouse: 20C day14C night
Final height of constant
temperature line after:
  Final height of
continuously in
fluctuating temp.
constant temp. line x 100
1 generation in fluctuating temp.  30.32 ± .93* 55.56 ± 1.0 55%
2 generations in fluctuating temp. 38.69 ± .42  42.86 ± .58 90%
3 generations in fluctuating temp. 30.61 ± .79  32.30 ± .60 95%
b) Assay in constant temperature: 10C
Final height of constant
temperature line after:
  Control †  
1 generation in fluctuating temp. 27.33 ± .77* 36.09 ± .11 77%
2 generations in fluctuating temp. 17.7  ± .34* 20.06 ± .45 88%
* Significant difference from corresponding control at the 1% level.
† Controls in this experiment are those plants arising from seed produced on plants grown continuously in the
controlled fluctuation in temperature of the greenhouse. No previous growth in constant temperature.

Fig. 2. Schematic diagram showing the manner in which the experiments were conducted. Percentage figures indicate the ratio of height of the "treated" plants to that of the "controls:" "Controls" in these experiments are plants grown in a greenhouse with diurnal fluctuation in temperature. Asterisk indicates significant difference at the 1 per cent level.

TABLE 2. Influence of seed source on growth rate (mm./day increase in length
during linear phase of growth) of two varieties of peas grown at different constant
temperatures and 16 hr. artificial light (approximately 1000 ft.-c.) per day

A: Variety 'L5'      
    Growing Temperature
                Seed Source 10C 14C 20C
  Greenhouse: ** ** **
       23C Day
     17C Night
8.37 ± 0.24 7.72 ± 0.24 6.22 ± 0.12
       20C Day
     14C Night
7.35 ± 0.22 8.57 ± 0.11 4.85 ± 0.16
B: Variety 'Unica'      
    Growing Temperature
  Seed Source 10C 14C 20C
  Field grown   **  
       (Holland)   7.31 ± 0.52 11.81 ± 0.18
  Greenhouse: ** ** **
       23C Day
     17C Night
5.20 ± 0.16 6.93-0.43 9.38 ± 0.36
  Greenhouse: ** ** **
       20C Day
     14C Night
4.85 ± 0.11 5.71 ± 0.16 10.90 ± 0.39
  Greenhouse: **    
       17C Day
     11C Night
8.35 ± 0.20    
**Difference between means significant at 1% level ("t" test).
*Difference between means significant at 5% level ("t" test) 16-18 plants/treatment.

Quantitative variation in a pure line as a function of previous history.Growth studies in a constant temperature environment were made on offspring derived from parents which had grown under different fluctuating temperature patterns. Table 2 shows the results of growing seeds from different diurnally fluctuating greenhouses in a constant temperature. The results for both varieties show an effect of seed source on the growth rates of the plants when grown at a constant temperature.

Fig. 3. The effects of a constant temperature of 10C. on the growth and development of peas as affected by the number of generations the plants were grown at this temperature.

Further studies were made with seeds descended from plants grown for 4-5 generations in three greenhouses, i.e., 17C. day, 11C. night; 20C. day, 14C. night; and 23C. day, 17C. night. The seeds were divided into three groups and each group was grown in one of the three greenhouses. The results are shown in table 3. Again, it can be seen that the parental environment affects the behavior of the offspring.

DISCUSSION.The experiments on the effects of a constant temperature show that such an environment is detrimental to the general growth and development of pea plants. That this effect is not unique to peas has been amply demonstrated by work carried out in the Earhart Laboratory. Went (1953) showed the requirement for a thermoperiod in tomato, especially with respect to fruit set, but also with respect to vegetative development. Hellmers (unpublished) has also shown that constant temperature is inhibitory, as has Kramer (1957). The present experiments, however, show that the inhibition becomes more pronounced in successive generations. Consequently, the effect of constant temperature must be more than a mere inhibition in growth, since the effects are cumulative until the fourth or fifth generation when the inhibition levels off.

The low level of activity reached in the later generations of plants grown at a constant temperature is not immediately reversible. That is, plants which have been raised for several generations under the detrimental constant temperatures will give rise, on return to fluctuating temperatures, to inferior plants as compared with controls grown for the preceding generations in the more favorable fluctuating temperatures. This fact in itself may not be surprising since one could expect that seed produced in an unfavorable environment might be less vigorous than those produced in a favorable environment. What is surprising is the fact that there appears to be an hysteresis phenomenon. When seeds from the less vigorous plants are grown for two generations in fluctuating temperatures, they do not differ phenotypically from the control plants which have been grown for many generations in fluctuating temperatures. When seeds from the originally less vigorous plants are reintroduced into the constant temperature, the difference becomes immediately apparent, despite a two-generation sojourn in the favorable fluctuating conditions. The fact that this difference persists for at least three generations, although decreasing somewhat each generation, indicates a strong analogy with the phenomena of "Dauermodifikation" as described by Jollos (1921).

TABLE 3. All plants grown at indicated fluctuating temperatures in natural light supplemented
with artificial light to give 16-hr. photoperiod. Daily increase in Ht. (mm.)





  20C. Day14C. Night 23°C. Day17C. Night
Seed Source 'L5' 'Unica'   'Unica'
Greenhouse: ** ** ** 5*
     23C. Day
     17 C. Night
18.84 ± .23 9.56 ± .14 6.75 ± .18 3.03 ± .08
Greenhouse: S5* **      
     20C. Day
     14° C. Night
17.19 ± .73      
Greenhouse: ** S4 S4 S4 *
     17C. Day
     11°C. Night
15.47 ± .50 8.91 ± .13 5.74 ± .18 3.26 ± .08
Field grown       **
     (Holland)       3.78 ± .09
Sn*-number of selfed generations at given temperatures. 16 to 18 plants in each treatment.
** Significant diff. at 1% level.

As with the constant temperature experiments, a continuous exposure to a specific thermoperiod was shown to exert a strong effect on the plants and their offspring (table 2). The temperature patterns used were such that the plants grown in each one could only be described as healthy, vigorous ones, yet the effect of the previous environmental history is apparent in the general growth behavior of the plants.

The results of these experiments show that the phenotype of the plant is the summation of its genetic heritage, together with the resultant or weighted summation of the parent and present environments. Perhaps one could envisage the weighting factor of any environmental effect as being inversely proportional to the number of generations by which that environment is removed from the generations under consideration.


Pea (Pisum sativum) plants were grown under conditions of constant temperature and artificial light or in controlled-temperature greenhouses in which there was a diurnally fluctuating temperature. It was found that: (1) A constant temperature is inhibitory for growth despite the fact that it may be the most nearly optimal constant temperature; (2) the inhibitory effect of a continuous exposure to constant temperature is cumulative from generation to generation reaching saturation with approximately the fifth generation; (3) differences in the temperature conditions of growth, whether constant or diurnally fluctuating, result in differences in the growth characteristics of the offspring of plants grown in these controlled conditions. The results of these experiments show that the phenotype of the plant is the summation of its genetic heritage together with the summation of the parent and present environment. Any environmental effect is inversely proportional to the number of generations by which that environment is removed from the generations under consideration.