The increasing computing power and advancement of image analysis software in the last decade made it possible to use high-resolution time-lapse movies for growth phenotyping in leaves and roots [28, 29]. Taking advantage of this technical advancement and bringing it further to integrate leaf and root growth analysis for high-resolution whole-plant phenotyping (Figure 1, Additional file 1: Figure S1), we have investigated the effects of root cooling and root illumination on diel growth patterns of roots and leaves in tobacco seedlings. While root growth responded to the root cooling and root illumination (Figures 2b,d,f; 3; Additional file 2: Figure S2), the overall diel leaf RGR pattern and the average leaf RGR values were not affected in the seedlings by the root-zone treatments (Figures 2a,c,e; 4; Additional file 3: Figure S3). These results, obtained simultaneously in growing roots and leaves of the same plants, corroborate the sensitivity of root growth velocity to temporal changes in environments [9, 12], in striking contrast to the robust diel growth patterns maintained in leaves.
When environmental conditions are kept constant, root growth remains relatively stable in a number of species, including N. tabacum[9, 12, 18, 19]. Our experiments confirmed that root growth velocity in N. tabacum is stable under constant conditions, such as shown in the control and the root cooling treatment (Figure 2b,d; Additional file 2: Figure S2). Yet, a clear decrease in root growth velocity was seen when the root temperature was lowered to 10°C (Figure 3a; Additional file 2: Figure S2). Low root-zone temperature reduces water uptake and root hydraulic conductivity [30, 31], and is furthermore known to restrict growth processes involved in apical root elongation [32, 33]. A negative effect on shoot growth is also expected as root cooling can impede water, nutrient or hormone supply to the shoot [34–36]. In fact, low root-zone temperature can decrease shoot biomass production in A. thaliana even at a normal (non-cooled) air temperature . Contrary to this expectation, we observed no reduction in leaf RGR of N. tabacum seedlings in the root cooling treatment (Figures 2a,c, 4a; Additional file 3: Figure S3). Davies & Van Volkenburgh  showed that low root temperature inhibits leaf growth of Phaseolus trifoliates primarily during the day but not at night. Recently, it has also been demonstrated that biomass accumulation and specific leaf area are more influenced by day temperature than night temperature in A. thaliana. Since the major leaf growth occurred predominantly during the night in N. tabacum seedlings (Figures 2a,c,e; 4b), negative effects of low root temperature on leaf RGR may not have been that pronounced in our experiments.
Large transient fluctuations in leaf RGR are often observed when illumination is switched on abruptly (Figure 2a-c; Additional file 3: Figure S3). They are partially attributed to changes in cell turgor – the driving force of cell expansion – caused by a sudden increase in evapotranspirational demand upon an increase in light intensity and stomatal conductance. Our synchronous leaf and root growth analysis detected similar but much lesser changes in root growth velocity concomitant with the leaf RGR fluctuations in the control and root cooling plants at dawn (Figure 2b,d), suggesting a momentary limitation to root growth during a transient increase of evapotranspirational demand in the leaves.
In contrast, when both shoot and root were exposed to the light-on event of the LD cycles, the transient changes in leaf RGR were somewhat suppressed and the root growth velocity remained low for several hours after a decrease at dawn (Figure 2e,f). This root growth repression in the morning can be attributed to the root illumination which is known to have an inhibitory effect on root growth [22, 23]. As was the case in the root cooling treatment, the reduced root growth by the root illumination (Figure 3a) did not obviously affect overall leaf growth rates and diel pattern in the tobacco seedlings (Figure 4). The strong decrease in root growth velocity without any change in leaf growth (i.e. net decrease in the whole-plant growth) suggests an impaired carbon and energy acquisition and/or altered carbon allocation in the plants under these two conditions. With respect to the diel fluctuation of root growth, it seems that the morning repression following the light-on event, together with the short but similarly large decrease upon the light-off event at the beginning of the dark period, gives rise to diel growth rhythmicity in the root system exposed to LD cycles (Figure 2f). Consistent with this, marked diel oscillations of root tip growth, albeit not quite the same patterns as found in the N. tabacum seedlings in this study, have been reported in the experiments with A. thaliana, in which the hypocotyl and the entire root system were subjected to LD cycles . Interestingly, the Arabidopsis circadian mutant elf3 did not show any oscillation in root growth under the whole-plant LD conditions . The peculiar root growth of elf3, together with the altered sensitivity of root growth to light exposure (and also to gravity) seen in several different clock mutants , points to an interaction between the circadian clock (intrinsic control) and light (environment) in the observed root growth inhibition.
Previous studies have established that leaf expansion in dicot species has a diel rhythmicity [7, 8] that is maintained under constant light and temperature conditions  presumably by the circadian clock . Most diel growth patterns observed in dicot leaves, including N. tabacum, are sinusoidal in nature and the growth rates reach the maximum amplitude at around dusk or dawn . In the N. tabacum seedlings studied here, the diel leaf RGR pattern did not exhibit sinusoidal oscillations but was instead characterised by high and low stable leaf RGR at night and during the day, respectively (Figure 2; Additional file 3: Figure S3), which deviates from the Type 1 growth pattern typically found in leaves of N. tabacum at a later developmental stage [1, 19]. The leaf growth patterns of N. tabacum reported in the previous studies of our group [1, 19] were obtained by using similar imaging and processing methods as the ones described here. Therefore, the distinct leaf growth pattern observed in the seedlings in the present work cannot be due to technical or methodological influence, but might indicate a developmental transition between leaves of seedling plants (used here) and leaves of older plants (used in the earlier literature). In fact, relatively stable RGR with small diel amplitudes has been reported in leaf 4 of young N. tabacum plants . Seedlings that have just passed the hypocotyl stage and have formed the first one or two primary leaves are in a developmental transition: from the hypocotyl stage in which they entirely rely on the stored resources to the seedling stage in which they become more and more dependent on their environment for resources (light, CO2, minerals and water). We speculate that the observed diel leaf RGR pattern, which does not follow the Type 1 growth pattern described in older plants of N. tabacum, may be specific to the early seedling stage. Such developmental influence on diel leaf growth is also supported by the earlier observations in poplar leaves , showing an increase of diel growth amplitudes with progressing season (low amplitudes in early summer, high amplitudes towards autumn), although this was not accompanied by a substantial change in the timing of maxima and minima of the diel leaf growth pattern.