Full Thesis is on Research Gate DOI:10.13140/RG.2.2.14708.63368
CHAPTER 4
Characterisation of Stomatal Development Following Transfer of Rice From One Light Environment to Another
4.1 Introduction
In Chapter 3 it was shown that rice leaves continuously grown in HL conditions have characteristics typical of sun leaves, for example increased thickness. One of the observations made was that stomata (controllable pores on the leaf epidermis) showed distinct patterns of size depending on the irradiance condition the plant was grown under, with high irradiance leading to larger stomata. This raises the question of when during leaf development stomatal size parameters are set and whether there is a sensitivity window during which environmental conditions, such as irradiance, can influence these parameters.
As outlined in Chapter 1, stomata are the result of patterned cell divisions in the epidermis of leaves. Cell division decreases during leaf development in a progressive manner so that processes dependent on cell division (such as stomatal formation) cease much earlier than growth of the leaf (i.e., mature leaves cannot adjust their number of stomata) (Fleming, 2005). This situation is further complicated in grass leaves where their unique cellular organisation means that stomatal differentiation is restricted to particular files of cells. File formation is also controlled by cell division events and the width of stomata will be set by the width of the cell files formed. Stomata size (defined by length and width) will have a strong influence on maximal pore aperture, thus potentially limiting the maximum gas exchange that can occur (Franks and Beerling, 2009). Thus, cell division events during early leaf differentiation will likely define the size and number of mature stomata, determining the limits of stomatal function in gas exchange in mature leaves. Exactly when the events of stomatal cell division terminate in rice, their responsiveness to irradiance level, and to what extent this influences the photosynthetic performance of mature leaves is the focus of this chapter.
Previous work in this area has shown that the ability of plants to acclimate to light is species dependant. For example stomatal density in Arabidopsis was reported to be higher in HL than LL conditions (Coupe et al., 2006) but remained approximately the same in rice (Hubbart et al., 2013). It has been demonstrated that the setting of stomatal parameters in young, developing leaves depends on signals from more mature leaves (Lake et al., 2001) and that this occurs in a range of plants (Coupe et al., 2005). With respect to rice, it has been demonstrated that the ability of a leaf to respond to altered irradiance with respect to photosynthetic acclimation becomes more limited after the leaf has emerged from the sheath of the surrounding, older leaf (Murchie et al., 2005). Overall, published work suggests that stomatal size might be set at a relatively early stage of rice leaf development, but the temporal resolution of previous studies have not allowed a precise resolution of this developmental response window.
Due to the regular generation of leaves by the shoot apical meristem, the time interval between the generation of successive leaves is highly predictable (under controlled growth conditions) and is termed the leaf plastochron (Erickson and Michelini, 1957). A convenient and robust way of defining the developmental stage of any leaf is to define its plastochron stage (P). During the initial events of development as the leaf primordium grows out of the meristem the leaf is defined as being at P1 stage. As soon as the subsequent leaf forms at the meristem, the leaf is defined as entering the P2 stage of development (with the younger leaf being at the P1 stage). Thus, for any plant defining a leaf by its order of formation (leaf 1, leaf 2, leaf 3, and so on) and its P stage (P1, P2, P3 and so on) provides a robust means of identifying and comparing developmentally equivalent leaves. Previous work in our group had already established such a plastochron staging system for rice (Van Campen et al., 2016). A particular advantage of this system in grasses, such as rice, is that it allows the prediction of the developmental stages of very young leaves hidden from view within the surrounding sheaths of older leaves, since (under controlled growth conditions) there is a strong predictive correlation of the P stage of older leaves with the P stage of younger leaves. In the experiments described in this chapter, this plastochron staging system was used to switch rice plants between LL and HL conditions so that leaves at early stages of development (P1, P2, P3, P4) differentiated under different irradiance levels during different phases of their development, allowing me to investigate the outcome on stomatal size and density in the mature leaves. This allowed me to investigate questions such as: if the early stage of a leaf’s development occurs in a plant under LL but the later stage under HL, are the stomatal parameters typical of leaves grown continuously under LL or HL (or intermediate)? At which P stage does the leaf lose its ability to respond to altered external environment?
As highlighted in previous chapters, stomata act as gas exchange regulators for the leaf. By modulating stomatal conductance (gs) in response to environmental stimuli, they can optimize carbon uptake for photosynthesis while minimizing excessive water loss from transpiration (Farquhar et al., 1980; Kim et al., 2010). Maximal stomatal conductance occurs when stomata are opened to their widest possible aperture and is determined by two physical properties of stomata, namely size (guard cell length multiplied by guard cell pair width) and density (stomata number per unit area) (Franks et al., 2009; Paul et al., 2012). Smaller stomata have the advantage of a faster response time to altered environment (Hetherington and Woodward, 2003; Franks and Farquhar, 2007), thus when coupled with high stomatal density may enable a leaf to achieve high stomatal conductance when plants are in favourable growth conditions and rapidly reduce conductance when conditions are unfavourable (Drake et al., 2013). However, Bussis et al. (2006) have shown that increased stomatal density can be compensated by reduced stomatal aperture, leading to the leaf maintaining a constant ratio of internal CO2 concentration (Ci) to ambient CO2 concentration (Ca). Thus, changes in stomatal size and density do not necessarily lead to changes in photosynthesis since plants display a high degree of plasticity in their ability to compensate for altered environmental conditions. Moreover, these differences can be species specific. For example, transfer from LL to HL conditions significantly increased photosynthesis in Betula ermanii and Acer rufinerve but not in Fagus crenata and Alocasia maccrorhiza (Sims and Pearcy, 1991; Oguchi et al., 2005). Overall, the relationship between irradiance level and photosynthesis is complicated and stomatal properties represent only one parameter in this relationship. Thus, the extent to which stomatal properties in rice determine or limit photosynthetic performance remains an open question.
The experiments described in this chapter aimed to define the developmental window for stomatal size response to altered irradiance, therefore we expected to generate rice leaves with a range of stomatal parameters. We therefore extended these experiments to include a series of gas exchange analyses on the mature leaves to examine the outcome of any changes in stomatal properties on photosynthetic performance. To what extent do leaves possess mechanisms that might maintain photosynthetic performance despite variation in stomatal properties arising during leaf development as a result of variation in the environment?
To summarise, in this chapter a series of experiments designed to investigate the influence of the light environment on stomata differentiation at different stages of rice leaf development are reported. Leaves were grown under one irradiance (high light, HL or low light, LL) then transferred to the alternative irradiance (HL to LL or HL to LL) at specific plastochron stages of leaf development (P1 to P5). The leaves were then allowed to grow to maturity and a series of measurements of stomatal size and leaf epidermal characteristics were made. Selected leaves were also analysed for various parameters of photosynthesis to investigate the outcome of any altered stomatal characteristics on these processes.
It should be noted that a portion of this chapter has already been published (van Campen et al., 2016), establishing that the P3-P4 stage transition of the developing rice leaf is the latest phase of a developmental window during which the leaf can respond to altered plant irradiance with respect changes in stomatal properties.
4.2 Aims
1. Identify the stages of rice leaf development during which altered irradiance can influence stomatal patterning.
2. Investigate the outcome of developmental-stage specific altered stomatal patterning on rice leaf photosynthetic performance
4.3 Brief Methodology
All rice plants were grown hydroponically using the growth chamber settings and nutrient solution (described in section 2.1). The plants were grown in either HL or LL conditions (section 2.1) and when they had reached a particular plastochron stage (P1, P3 or P5) they were transferred to the opposite light condition. Since the developing leaf primordia were concealed in layers of leaf sheaths, the length of leaf no. 3 was used as a proxy to approximate leaf no. 5 P-stage (described in 2.2). For plants transferred from HL to LL, leaf 5 was sampled and measured 5 - 6 days after transfer (depending upon the plastochron stage when transfer took place), while for LL to HL transfer the sampling and measurements were made between 3 - 4 days after transfer. Two identical transfer identical experiments were carried out, one to measure stomatal and leaf development characteristics and the other to measure physiological and biochemical leaf parameters.
4.3.1 Stomatal and Epidermal Measurements
Rice leaves were prepared using the methods described in sections 2.4, 2.5 and 2.6.1. Measurements of SD, SCA and SPA, SCW, SCL, GCW, and the percentage of files containing stomata were made following the methods described in section 2.6.2. For each parameter five measurements per leaf were taken and averaged to obtain one data point. Eleven replicate leaves were used.
4.3.2 Physiological Measurements and Biochemical Analyses
Physiological and biochemical measurements were again made on the middle section of the fully expanded leaf 5. Three replicate leaves were used. Gas exchange measurements were carried out to obtain A-Ci curves as described in section 2.3.1. PAR was maintained at 2000 µmol m–2s–1 with 10% blue light and 90% red light. Stomatal conductance versus intercellular CO2 concentration curves (gs-Ci) were also obtained. Intrinsic water use efficiency (iWUE) was computed by taking the ratio between assimilation and stomatal conductance at ambient (400 ppm) CO2 (Ca) as well as when iWUE was at its maximum (iWUEmax), which was always at the last point of A-Ci and gs-Ci curves.
Pigment quantification and isotopic carbon discrimination determination were performed as described in sections 2.3.2 and 2.3.3 respectively using five replicate leaves for each analysis.
4.3.3 Statistical Analysis
For comparison of means, analysis of variance (ANOVA) was employed but when data were non-parametric (according to D'Agostino & Pearson omnibus normality test), Kruskal Wallis multiple comparisons test (add-on macro) was used instead. The relationship between variables was assessed using Pearson’s correlation analysis where non-normal data were transformed using the best rounded λ (lambda) values suggested by Box-Cox transformation in Minitab. An F-test was also performed when the variance between two samples needed to be compared for similarity. All statistical analyses, graphs and contour plot were prepared using Microsoft Excel 2016, GraphPad Prism version 6 and Minitab 17 software.
4.4 Results
4.4.1 Effect of transfer of leaves from High Light to Low Light environment at different developmental stages on final stomatal parameters
Considering first Stomatal Complex Area (SCA), the mean value under continuous LL (262 µm2) was significantly lower than observed in mature leaf 5 maintained under continuous HL conditions (342 µm2) (Fig. 4.1A). On transfer from HL to LL at early stages in development (P1, P3) the final mean SCA was similar to those observed in leaves maintained continuously under LL. When plants were transferred at a later stage (P5), mean SCA was intermediate between HL and LL values.
One interesting observation was that although the variances measured in the HL, LL and P5 treatments were relatively small (Coefficient of variation, cv of 6.9%, 10.9% and 9.0% respectively), the values observed in the P3 and especially the P1 transfer experiments were much higher (cv of 13.5% and 26.5% respectively). A two-tailed F-test (α=0.05) showed that there was a significant difference in variance (p=0.01) between P1 and LL treatments but there was no significantly different in variance between P3 and HL treatments (p=0.14).
Measurements of SCA, SCW and SCL are shown in Fig. 4.1 A B and C respectively. Similar tendencies were observed for both SCW and SCL. Stomatal complexes under HL conditions were wider and longer than observed under LL conditions. After transfer at P1 and P3 stages SCW and SCL values were more similar to those measured in leaves kept continuously under LL. Once again, although the SCW variances measured for HL, LL and P5-transfer leaves were relatively small (cv of 7.3%, 5.2% and 5.7% respectively), for leaves transferred at P3 and especially P1 stage the variances measured were much higher (cv of 6.6% and 15.4% respectively).
Figure 4.1:
Measurements of stomatal size (A-E), epidermal size (F and H) and stomatal patterning (G and I) in mature L5 grown in HL condition that were transferred to LL condition at different P-stages, as indicated. ANOVA followed by a post hoc Tukey’s HSD test (av) was performed on data showing a normal distribution while the non-parametric Kruskal-Wallis multiple comparison test (kw) was performed on data showing non-normal distribution. Each black dot represents a single measurement. Vertical error bars represent standard error of means while the dotted horizontal red lines are the means. Treatments that share same letter in a grouping cannot be statistically distinguished (P-value <0.05).
Again, the P1 data had a significantly different variance (p<0.01) compared to the LL control but this was not the case for the P3 treatment data. The same trend in variance was also observed for SCL variance analysis, i.e., P1 and P3 transfer value showed a relatively high variance compared to the other data sets in the experiment. Correlation analysis also showed that both SCW and SCL showed a highly significant positive correlation with SCA (r= 0.72 and r=0.82 respectively, p< 0.01, Table 4.1).
Table 4.1:
Pearson’s correlation coefficients (r) among the pooled values of stomatal size, patterning and leaf epidermal properties from the transfer experiments (n=11 for each parameter). Single asterisk indicates correlations which are significant at p<0.05 confidence limit while double asterisks * indicate correlations which are significant at p<0.01 confidence limit.
Parameter | SCA | SCW | SCL | GCW | SPA | CFN | SF% | CFW | SD |
SCA |
|
|
|
|
|
|
|
| |
SCW | 0.72** |
|
|
|
|
|
|
|
|
SCL | 0.82** | 0.60** |
|
|
|
|
|
| |
GCW | 0.45** | 0.55** | 0.34** |
|
|
|
|
|
|
SPA | 0.81** | 0.46** | 0.73** | 0.37** |
|
|
|
| |
CFN | -0.16 | -0.24* | 0.00 | -0.23* | -0.14 |
|
|
|
|
SF% | 0.26* | 0.27* | 0.21* | 0.10 | 0.17 | -0.16 |
|
| |
CFW | 0.67** | 0.70** | 0.50** | 0.51** | 0.57** | -0.54** | 0.28** |
|
|
SD | -0.58** | -0.44** | -0.51** | -0.28** | -0.52** | 0.03 | 0.24* | -0.57** |
Each stomatal complex contains a guard cell pair which forms the actual stomatal pore. To investigate the extent to which measurements of stomatal complex dimension acted as a proxy for guard cell dimension and pore size, these parameters were analysed. As shown in Fig. 4.1D and 4.1E the pattern of guard cell width (GCW) and stomatal pore area (SPA) in leaves after the different treatments was similar. Highest mean values of GCW and SPA were measured in leaves grown continuously under HL (5.2µm and 90.6µm2 respectively) and the mean values for these parameters were lower in leaves grown continuously under LL (4.5µm and 63.0µm2 respectively). Leaves transferred at stage P1 from HL to LL showed relatively low means of GCW and SPA whereas leaves transferred at stages P3 and P5 showed mean values either similar to LL samples (for GCW) or intermediate between HL and LL samples (for SPA). GCW and SPA values showed a positive correlation with SCA (r= 0.45 and r=0.81 respectively, p< 0.01, Table 4.1).
Overall, the data shown in Fig. 4.1A-E suggested that leaves grown under LL had smaller stomatal complexes as well as other size related parameters than those grown under HL, and that transfer from HL to LL resulted in smaller stomatal complexes even when the transfer was performed relatively late in development (P5). For some parameters such as SCA, SCW and SCL, when the transfer was performed at relatively early stages of development (P3 and/or P1) there was a marked variation in the measured values, with some extremely small values observed even when compared to the LL leaves. The correlation analysis in Table 4.1 shows that SCA parameter is a good proxy to estimate other stomatal size related parameters as they all had a highly significant positive correlation with SCA.
To investigate what outcome these shifts in SCA had on stomatal density (SD), this parameter was measured in the same leaves samples analysed in Fig. 4.1A-E. The results of this analysis (shown in Fig. 4.1I) indicated that for leaves maintained under continual LL or HL the differences in the mean SD values were minimal (261 and 269 stomata mm-2 respectively). Although mean SD values in leaves transferred from HL to LL at different developmental stages (P1, P3, P5) tended to be slightly higher than those observed in HL and LL, analysis of the data revealed that the transferred leaves were characterised by a massive spread of SD values, with P1 transferred leaves showing SD value as high as 624 stomata mm-2 compared with extreme SD values in HL leaves of 362 stomata mm-2 and in LL leaves of 373 stomata mm-2. A similar trend in other size parameters was also observed, with variances measured in the HL, LL and HL-LL P5 samples generally being relatively small but large variance being observed when transfer was performed at P1 and P3 stages. Since unequal variance violates the requirements for ANOVA, a non-parametric Kruskal-Wallis test was used to analyse SD. By this analysis the mean SD did not significantly differ between any of the treatments, but this similarity in mean value obscured a difference in variance depending on leaf stage at transfer from HL to LL.
SD in a rice leaf is likely to reflect both the size of stomatal complexes and the number/size of cell files within which stomata arise. To investigate the potential relationship of stomatal density and size to these parameters epidermal file number within interveinal gaps and the number of cell files within which stomata arose were also studied. As shown in Fig. 4.1F leaves kept under continuous LL had a slightly higher mean number of cell files than HL leaves (about 14 cell files), but the spread of these values was higher under HL than LL conditions (cv of 8.7% and 4.7% respectively). In the P1 and P3-transfer experiments, leaves showed higher mean values for CFN, with P3 transferred leaves having an extreme value of 18 cell files and a minimum of 14 whereas HL leaves had a maximum of 16 cell files and a minimum of 12. Leaves kept under continuous LL had a maximum CFN of 15 and a minimum of 13. CFN showed a significant negative correlation with cell file width (CFW, r = -0.54, p< 0.01, Table 4.1) but when the number of cell files containing any stomata is considered (Fig. 4.3G) a very weak inverse relationship (r=-0.16) to CFN was observed. Thus for leaves maintained continuously under HL the percentage of stomata-containing cell files was significantly higher than for LL leaves and leaves transferred at P1 or P3 stage, suggesting that P3 is the last stage able to initiate additional rows of stomatal or epidermal files in response to altered irradiance. For any given interveinal gap sampled the maximum number of stomatal files was always four.
Since in rice leaves stomata arise in epidermal cell files the width of these files (CFW) is likely to influence the size of the stomatal complexes. To test this idea, the cell file width (CFW) in the transferred and control plants was measured. As shown in Fig. 4.3H there was a similar pattern of CFW and the stomatal complex width (SCW) in Fig. 4.3C (r = 0.70, Table 4.1) and a significant inverse correlation with cell file number (Fig. 4.3F) (r = -0.24). Thus, mean CFW was significantly high in leaves grown under continuous HL (13.1µm) and those transferred at P5 from HL to LL (12.3 µm), whereas CFW was significantly smaller in LL (11.4µm) leaves and those transferred from HL to LL at early stages of P1 and P3 (10.9µm and 11.3µm respectively). Once again, the P1 and P3 transferred leaves were characterised by relatively high variances for the parameter measured, with some extremely narrow cell files measured in the P1 transferred leaves (8.3µm).
4.4.2 Effect of transfer of leaves from Low Light to High Light environment at different developmental stages on stomatal parameters
The results described in the previous section described the outcome on stomatal parameters of transferring plants from HL to LL conditions at various stages of leaf development. I also investigated the outcome of the reverse situation where plants were transferred from LL to HL conditions. As shown in Fig. 4.2A, when leaves were transferred at very early stage of development namely P1, the mean value of SCA in the mature leaves was similar to that measured in plants grown continuously under HL. When the transfer occurred later in development (stage P3 and P5), mean SCA remained similar to that measured in plants grown under continuous LL. The variance in the values observed under transfer conditions especially P5 tended to be higher (cv of 17.3%) than those values observed in P1 and P3 transfers and plants maintained under continual HL or LL (cv of 11.4%, 11.4%, 10.9% and 6.9% respectively). To investigate the contribution of change in stomatal complex width (SCW) and length (SCL) to the observed changes in SCA, I measured these parameters in the various treatments. As shown in Fig. 4.2B and 4.2C, similar patterns of change in values were observed in both SCW and SCL, although the absolute value of these differences were often quite small. Thus, after transfer at very early stage of leaf development (P1), there were slight increases both in mean SCW and SCL whereas in the P3 and P5 transferred leaves changes in SCW and SCL were minimal. Correlation analysis in Table 4.1 has shown that there were significant positive relationships between SCA with SCW and SCL, thus showing these parameters are good predictors for each other.
Figure 4.2:
Measurements of stomatal size (A-E), epidermal size (F and H) and stomatal patterning (G and I) in mature L5 grown in LL condition that were transferred to HL condition at different P-stages, as indicated. ANOVA followed by a post hoc Tukey’s HSD test (av) was performed on data showing a normal distribution while the non-parametric Kruskal-Wallis multiple comparison test (kw) was performed on data showing non-normal distribution. Each black dot represents a single measurement. Vertical error bars represent standard error of means while the dotted horizontal red lines are the means. Treatments that share same letter in a grouping cannot be statistically distinguished (P-value <0.05).
With respect to guard cell width (GCW) and stomatal pore area (SPA), generally mean GCW showed no increase in leaves transferred at any developmental stage, in fact it slightly decreased under P3 and P5 transfers, whereas for plants maintained under continuous HL GCW was significantly higher than in LL-maintained or transferred leaves (Fig. 4.2D). In contrast, estimates of SPA (Fig. 4.2E) indicated that mean values in early transfer (P1) leaves were higher than those measured in P3 and P5 transfer, which were similar to HL leaves. The high variance measured for all of these samples meant that overall there was no significant difference in SPA.
In terms of altered stomatal density (SD), leaves grown under continuous HL and LL and under P1 and P5 transfers had similar SD (Fig. 4.2I). Surprisingly the transfer at P3 resulted in higher mean SD values than the rest of the treatments and this was mainly due to a few leaves in this transfer showing unusually very high densities above 400 stomata mm-2. To investigate whether the differences in SD observed in some transferred samples reflected differences in the number of cell files in the interveinal gap, values of cell file number (CFN), width (CFW) and the number of cell files containing stomata were also measured. With respect to CFN, although there were some differences in mean values for the different treatments, the mean CFN value (about 14) was comparable for all treatments (Fig. 4.2F), although it was noticeable that the P1 transferred samples had the most values of CFN 14 or greater. In terms of CFW, the P5 transferred samples were distinguished by having some extremely narrow cell files (8.4µm), leading to a significantly different CFW from leaves maintained under continual HL (Fig. 4.2H). CFN and CFW also showed a significant inverse correlation (r = -0.54, Table 4.1) thus can be used as a good predictor for one another. The percentage of cell files containing stomata in the P5 transferred leaves was similar to those maintained under continuous LL, whereas the P3 transferred leaves had mean values very similar to those measured in HL leaves (Fig. 4.2H). P1 transferred leaves however had intermediate values between HL and LL. There was, however, a high variation in percentage of files containing stomata in all treatments. Despite this, there was still a positive significant, although weak, relationship between files containing stomata and CFW (r = 0.28).
4.4.3 Summary of effect of transfer between high and low light environments at specific leaf developmental stages on stomatal pattern and size
With respect to the transfer either from HL to LL or LL to HL, the former in general allows significant changes in almost all the parameters after P1 transfer (Fig. 4.3A). Leaves transferred at the later P3 stage showed less ability in significantly changing as many parameters as the transfer at P1, whereas transfer at the later P5 stage did not lead to any significant change in any parameter. The high variation of SCA values, especially in P1 and P3 transfer treatments, implied developmental stages where stomatal phenotypes are highly plastic with respect to environmental signals such as the irradiance intensities used in this experiment. Differences in SCA measurements are determined by both length and width properties of the complex where SCL in general showed a higher degree of positive correlation with SCA. Due to the linear nature of stomatal arrangement in grass leaves, width of SCA is to some extent dependent on cell file width. CFW was most plastic at P1 and P3 stages but by P5 it becomes less responsive to altered irradiance. Similarly, the number of cell files formed (reflecting longitudinal cell divisions during leaf development) is flexible at P1 and P3 but much less flexible at P5 stage.
In rice leaves, it appears that some cell files are stomata-forming while others are not (i.e., stomatal patterning is set by cell file interactions), as evident in Fig. 2.7). It is worth noting that most interveinal gaps in rice leaves contain a maximum of four stomata-containing files while the total number of cell files may vary considerably. This may be connected with stomata position as stomata are always formed over mesophyll cells but not over bulliform cells. Leaves transferred at P1 and P3 stage tend to have many, narrow files but only a small percentage of these form stomata. By P5, the number and width of cell files is set, as well the percentage of cell files containing stomata.
Figure 4.3:
Summary of possible windows of sensitivity at different plastochron stages (P1 to P5) for the transfer from HL to LL (A, based on Fig. 4.1) and LL to HL (B, based on Fig. 4.2) for various stomatal and epidermal properties. Each tick (✔) symbol under any P-stage denotes the possibility of altering the corresponding parameter compared to the initial light condition (shaded column on the left).
The resultant stomatal patterning shows that there is little major difference between SD in HL and LL leaves but HL leaves have significantly more stomatal containing cell files than LL leaves. The other main difference is stomatal size (SCA), with HL leaves having larger stomata due to an increase in cell file width and the length of the cells formed within a file.
The above points describe the transfer events for transfer from HL to LL. For LL to HL transfer, however, the same ability to significantly change the studied parameters after P1 and P3 transfer was not as clear. This direction of transfer generally led to less significant alteration in the parameters measured (Fig. 4.3B). Changes are more minor, less pronounced or intermediate. Even at the early P1 and P3 stages the parameters measured seem less plastic and responsive than plants switched from HL to LL conditions. Although the transfer at P3 could significantly alter the SPA, the spread of the data (in Fig. 4.2E) were much larger compared to the transfer at P1 and P3. Nevertheless, it is interesting to note that LL to HL transfer still led to a significantly change in guard cell width at each P stage analysed.
4.4.4 Analysis of the outcome of altered stomatal pattern and size on leaf physiology
The results of the analyses in Fig. 4.1 and Fig. 4.2 indicated that by switching rice leaves between different light environments at particular stages of leaf development it was possible to generate leaves with distinct stomatal densities and size. To investigate the outcome of these changes on leaf performance, I performed a series of physiological analyses on some of these leaves (n=3). These leaves described in each subsection below dealing with photosynthetic assimilation, stomatal conductance and intrinsic water use efficiency provided paired data with the measurement of stomatal structure reported in the previous section. This leads to the second hypothesis of this study; that some stomatal or epidermal properties in rice could be a good indicator of important leaf physiological characteristics, thus reflecting overall leaf performance and efficiency.
4.4.5 Assimilation Rate
A series of A-Ci curves was obtained for three plants, each either maintained under continuous HL or LL conditions or transferred from HL to LL at P1, P3 or P5 stage before assessment of physiology at maturity. Measurements were made immediately upon leaf no. 5 maturity (P6) which was marked by the appearance of the collar (joint in between leaf blade and the sheath). From the leaf emergence to this point, the leaf blade had been exposed to the new light condition (before measurements took place) about 3-4 and 5-6 days for HL and LL conditions, respectively. Measurements were taken very close to full leaf extension, thus ameliorating any potential for reduction in photosynthesis as a consequence of ageing (Murchie et al., 2005).
HL grown leaves in general (Fig. 4.4iv-HL) had higher maximum assimilation rate (Amax, about 46 mmol CO2 m-2 s-1) than LL grown leaves (about 33 mmol CO2 m-2 s-1). The first phase (Rubiso limited) of the A-Ci curve for HL grown leaves had a steeper slope than LL grown leaves (Fig. 4.4iv-LL) indicating rapid assimilation of CO2 suggesting an abundance of Rubisco in the HL leaves. The first phase also ended at lower Ci concentration (about 213 ppm) for HL grown leaves than LL leaves (about 243 ppm), thus a quicker shift to the next phase of the curve (RuBP regeneration limited), marked by a subtle decrease in the curve slope. HL grown leaves had some variance in both A and Ci values measured, especially in the second phase of the curve, whereas LL grown leaves had the same variance trend only for the A values but not Ci values, probably due the lower overall CO2 assimilation compared to HL grown leaves enabling stomata to more rapidly adjust to maintain a more consistent Ci environment.
It is quite surprising to see that transfer at P1, P3 and P5 stage (Fig. 4.4iv-P1, P3, P5) on average produced similar A-Ci curves and that the Amax and Ci concentration at which the shift into the second phase of the curve occurred was similar to those leaves grown in continuous LL. Further investigation by deriving useful photosynthetic parameters support this claim; the rate for A400 (Fig 4.5 A, about 20-25 mmol CO2 m-2 s-1) and Amax (Fig. 4.5 B, about 30 mmol CO2 m-2 s-1) for transfer at P1, P3 and P5 were similar to LL leaves. Such trends were tightly bound to the Vcmax (Fig. 4.5C) which also had the same trends as the assimilation rates, where LL leaves had Vcmax value of about 40-50 mmol CO2 m-2s-1.
Figure 4.4
A-Ci curves in mature rice leaves for the transfer experiment from HL to LL. Rows (i), (ii) and (iii) represent individual curves (total n=3) for the corresponding treatment labelled at the top of each column. Row (iv) represents mean A- Ci curve for the particular treatment where the horizontal-vertical bars represent standard error of mean. The blue dotted lines are best fit curves for CO2 assimilation rate limited by the amount and activity of Rubisco (enzyme limited/RuBP saturated) while the green dotted lines are best fit curves for CO2 assimilation rate limited by RuBP regeneration (light limited/RuBP limited). Each A-Ci curve is marked with Ca=400ppm to approximately show assimilation rate at the current ambient CO2 level while the experiment took place.
The relative stomatal limitations (ls) for HL (about 20%) and LL (about 15%) leaves were similar to the Vcmax trend but the P5-transferred leaves showed exceptionally low ls (about 7%) thus partially explaining their relatively higher Vcmax values compared to P1 and P3 transferred leaves (Fig. 4.5D). Jmax for the transfer at P5 stage (Fig. 4.5E) had an intermediate value between HL (closer) and LL leaves (about 125 mmol e- m-2s-1). Pigments analysis for chlorophyll a/b ratio (Fig. 4.5 F) also showed that the P5-transferred leaves had similar ratio (about 4) with HL leaves. thus further explaining their ability to have comparable A400 value with HL leaves. Transfer under all stages produced minimum variation with regard to assimilation rate but on average transfer at P3 produced the largest variation in Ci values, which were similar to HL grown leaves.
When the transfer was performed in the opposite direction from LL to HL condition, leaves transferred at P1 on average (Fig. 4.6iv-P1) had comparable A-Ci curve pattern to HL grown leaves. A400 analysis (Fig. 4.7A) showed that there was a tendency for HL and P1-transferred leaves to have higher A400 (both about 33 mmol CO2 m-2 s-1) values than LL leaves while transfer at later P3 and P5 stages resulted in similar A400 to LL leaves (about 18-25 mmol CO2 m-2 s-1). But the same trend was not evident in the Amax analysis (Fig. 4.7B) since transfer at any P-stages would result in comparable values to HL leaves (about 45 mmol CO2 m-2 s-1).
Figure 4.5
Mean assimilation rates measured at 400ppm CO2 (ambient level, A400) (A) and maximum assimilation rates (Amax) (B) achievable for the HL to LL transfer experiment. The measured photosynthetic rates are explained by similar trend in maximum carboxylation (Vcmax, C) whose capacity depends on CO2 supply influenced by the relative stomatal limitation (ls in D). The electron transport (Jmax, E) rates also have similar trend to assimilation and the capacity is explained by the ratio of chlorophyll a/b (F). All means are extracted from Fig. 4.4iv. One way ANOVA followed by a post hoc Tukey’s HSD test where n=3 with groups sharing different letters are significantly different (p< 0.05) and error bars represent standard error of mean.
Figure 4.6
A-Ci curves in mature rice leaves for the transfer experiment from LL to HL. Rows (i), (ii) and (iii) represent individual curves (total n=3) for the corresponding treatment labelled at the top of each column. Row (iv) represents mean A-Ci curve for the particular treatment where the horizontal-vertical bars represent standard error of mean. The blue dotted lines are best fit curves for CO2 assimilation rate limited by the amount and activity of Rubisco (enzyme limited/RuBP saturated) while the green dotted lines are best fit curves for CO2 assimilation rate limited by RuBP regeneration (light limited/RuBP limited). Each A-Ci curve is marked with Ca=400ppm to approximately show assimilation rate at the current ambient CO2 level while the experiment took place.
Transfer at later P3 and P5 stages decreased the Amax value to around 40 mmol CO2 m-2 s-1 but still much higher than those of LL grown leaves. Interestingly variations seen in P1 transferred leaves were also similar to those of HL grown leaves, while for transfer at P3 and P5 the variations were more pronounced in Ci values, especially in the second phase of the curve. These Ci variations could be explained by some odd individual curves (Fig.4.6iii-P3 and 4.6i-P5) where they looked steep in the first phase of the curve with a short lived second phase, suggesting rapid CO2 influx regulation as well as high assimilation rate. Vcmax analysis (Fig. 4.7C) showed that transfer at any P-stages would result in Vcmax value similar to HL leaves (80 mmol CO2 m-2s-1) where P1-transferred leaves had exceptionally high Vcmax value among all treatments of about 85 mmol CO2 m-2s-1. This indicates a comparable amount of Rubisco between HL and P1-transferred leaves as well as less obstructed CO2 diffusion, which was evident by a relatively lower (similar to LL leaves) stomatal limitation of about 15% (Fig. 4.7D). Jmax had the same trend as the Amax analysis earlier (Fig. 4.7 E) where HL leaves and the P-transferred leaves had comparable Jmax values around 200-250 mmol e- m-2s-1. Interestingly, all P-transferred leaves had similar chlorophyll a/b ratio (about 3.5, Fig. 4.7F) while LL grown leaves had the ratio of 3, indicating less investment in making chlorophyll-b for LHCII to capture more photons.
Figure 4.7
Mean assimilation rates measured at 400ppm CO2 (ambient level, A400) (A) and maximum assimilation rates (Amax) (B) achievable for the LL to HL transfer experiment. The measured photosynthetic rates are explained by similar trend in maximum carboxylation (Vcmax, C) whose capacity depends on CO2 supply influenced by the relative stomatal limitation (ls in D). The electron transport (Jmax, E) rates also have similar trend to assimilation and the capacity is explained by the ratio of chlorophyll a/b (F). All means are extracted from Fig. 4.4iv. One way ANOVA followed by a post hoc Tukey’s HSD test where n=3 with groups sharing different letters are significantly different (p< 0.05) and error bars represent standard error of mean.
Overall high light grown leaves undoubtedly had higher assimilation rates than LL grown leaves either measured under ambient CO2 level (A400) (Fig. 4.5A and 4.7A) or Amax (Fig. 4.5B and 4.7B) in both HL-LL-HL transfer directions. It is interesting to note that the rice leaf seems to have the ability to achieve Amax similar to the final light condition after transfer at any P-stage whereas A400 in general seems to follow the trend where transfer at P1 and/or P3 is the cut-off point to produce leaves with an assimilation rate similar to leaves continuously grown in the final light condition to which they are exposed.
4.4.6 Stomatal Conductance (gs)
A series of gs-Ci curves were also obtained from the three plants used in the previous subsection dealing with assimilation. There was a general trend for both HL and LL grown leaves (Fig. 4.8iv-HL and LL) where gs values increased steadily in low CO2 conditions before decreasing rapidly (HL leaves) or gradually (LL leaves) before reaching the lowest points. The key difference in overall curve shape was that HL grown leaves covered a wider range of gs with changing CO2 concentration, indicating a more rapid stomatal movement than LL grown leaves, which had steadier gs values and less variation. The transfer at P1 and P3 resulted in similar overall curve pattern to LL grown leaves but, surprisingly, P5 on average (Fig. 4.8iv-P5) had the highest gs values at any given Ci points. Fig. 4.9A shows mean gs values at 400ppm CO2 concentration (gs400) for the individual plants while Fig.4.8B shows the mean maximum gs values (gsmax) and for the same treatments. Interestingly in most cases gsmax values occurred before ambient CO2 level (marked as Ca=400ppm in Fig. 4.8). HL grown leaves on average had higher gs400 and gsmax than LL grown leaves but the leaves of P5 transfer unexpectedly had the highest overall gs values (Fig. 4.9A and 4.9B respectively).
Figure 4.8
gs-Ci curves in mature rice leaves for the transfer experiment from HL to LL. Rows (i), (ii) and (iii) represent individual curves (total n=3) for the corresponding treatment labelled at the top of each column. Row (iv) represents mean gs-Ci curve for the particular treatment where the horizontal-vertical bars represent standard error of mean. Each gs-Ci curve is marked with Ca=400ppm to approximately show assimilation rate at the current ambient CO2 level while the experiment took place.
Figure 4.9
Mean stomatal conductance (gs) rates measured at 400ppm CO2 (ambient level, gs400) (A) and maximum gs (gsmax) rate (B) achievable for the HL to LL transfer experiment. The means for both gs conditions were extracted from Fig. 4.8iv. One way ANOVA followed by a post hoc Tukey’s HSD test where n=3 with groups sharing different letters are significantly different (p< 0.05) and error bars represent standard error of mean.
In the opposite transfer from LL to HL conditions, the P1 transferred leaves unexpectedly produced a similar average curve to HL grown leaves (Fig. 4.10iv-P1). The later transfer at P5 stage also produced similar average curve to LL grown leaves, indicating that overall gs values followed the light condition to which the leaf had been the longest exposed. Mean gs400 and gsmax values showed high variation in the transfer treatments (Fig. 4.11A and 4.11B respectively). On average, gs400 and gsmax of the P1 (earliest) and P5 (latest) transfers achieved values similar to those measured in HL grown leaves.
Figure 4.10
gs-Ci curves in mature rice leaves for the transfer experiment from LL to HL. Rows (i), (ii) and (iii) represent individual curves (total n=3) for the corresponding treatment labelled at the top of each column. Row (iv) represents mean gs-Ci curve for the particular treatment where the horizontal-vertical bars represent standard error of mean. Each gs-Ci curve is marked with Ca=400ppm to approximately show assimilation rate at the current ambient CO2 level while the experiment took place.
Figure 4.11
Mean stomatal conductance (gs) rates measured at 400ppm CO2 (ambient level, gs400) (A) and maximum gs (gsmax) rate (B) achievable for the LL to HL transfer experiment. The means for both gs conditions were extracted from Fig. 4.10iv. One way ANOVA followed by a post hoc Tukey’s HSD test where n=3 with groups sharing different letters are significantly different (p< 0.05) and error bars represent standard error of mean.
However the transfer at P3 stage produced a perplexing dataset where the curve had a very narrow gs range (Fig. 4.10iii-P3) while the average gs400 and gsmax values were very low (Fig. 4.8C and 4.8D). Cross-referencing back to Fig 4.2 for the stomatal properties for these individual leaves does not indicate any unusual pattern in SCA size (Fig. 4.2A) and, in fact, SD on average was highest in this P3 transfer. The only plausible explanation for these irregular low gs values was that the stomata were relatively closed in this particular treatment
4.4.7 Intrinsic Water Use Efficiency (iWUE)
Results from the previous two subsections were brought together to calculate the iWUE for the leaves transferred between different irradiances at particular stages of development. iWUE is defined as the ratio of CO2 assimilation and stomatal conductance (A/gs). At 400ppm CO2 level (iWUE400), HL grown leaves in general had significantly higher iWUE (value about 100) than LL grown leaves (value about 40) (Fig. 4.12A), which had similar iWUE values as leaves transferred from HL to LL at P1, P3 and P5 stages. The transfer from HL to LL in general produced minimum variation in iWUE across all treatments. As the supply of CO2 increased so did the iWUE due to the proportional drop in gs values, indicating stomatal closure. It is worth to note that the calculation revealed the last point of A-Ci and gs-Ci curves always produced the highest iWUE (iWUEmax). HL grown leaves on average (value about 150) still had higher iWUEmax than LL grown leaves but surprisingly transfer at P3 stage had the highest average iWUEmax (value about 200) (Fig. 4.12B). This anomaly was due to the iWUEmax value of in one of the plants which showed a 368% increase from the iWUE400 value. This was caused by an extremely low value of gsmax, which in turn corresponded to an extremely low stomatal density in Fig. 4.1I.
In the reverse transfer experiment from LL to HL condition, iWUE400 values showed the same trend as seen in the other transfer mode where LL grown leaves had similar iWUE400 with the P1, P3 and P5 transfer treatments, while HL had the highest iWUE400 (Fig. 4.12C). There were high variations in iWUEmax for the transfer experiments (Fig. 4.12D) but in general transfer at P1 and P3 stage produced iWUEmax values most similar to HL grown leaves, while P5 transfer values were more similar to LL grown leaves (Fig. 4.12D). The occasional extremely high iWUEmax values measured for certain leaves were mainly caused by extremely low gs values.
Figure 4.12
Mean intrinsic water use efficiency (iWUE) rates measured at 400ppm CO2 (ambient level, iWUE400) for the HL-LL (A) and LL-HL (C) transfer experiments. The maximum iWUE (iwuemax) for the HL-LL (B) and LL-HL (D) transfer experiments are also calculated. iWUE is the ratio between CO2 assimilation and stomatal conductance (A/gs). Carbon isotope ratio (δ13C) is presented as per mill (‰) basis in (E) for HL-LL and (F) for LL-HL transfer experiments to validate the iWUE analyses. One way ANOVA followed by a post hoc Tukey’s HSD test where n=3 with groups sharing different letters are significantly different (p< 0.05) and error bars represent standard error of mean.
Carbon isotope discrimination analysis (δ13C) again showed that (as in Chapter 3) HL leaves had a more positive value (about -31.3‰) compared to LL leaves (about -33.3‰) (Fig. 4.12 E and F). The same trend was observed in δ13C values for the HL-LL transfer as seen in iWUE400 The inverse LL-HL transfer revealed a generally similar trend for iWUE400 and δ13C (Fig. 4.12 C and F respectively), although it was unusual that the P5-transferred leaves to had almost the same δ13C as the HL leaves since the iWUE400 value for such leaves was much lower (i.e., similar to LL leaves). This may be explained by the gs values in these P5 leaves being similar to HL leaves but Vcmax values being low compared to HL leaves.
4.4.8 Correlation Between Stomatal and Physiological Properties
The previous sections that dealt individually with assimilation, stomatal conductance and water use efficiency all came from the same plants. In addition to these physio-biochemical analyses, the leaves samples from the same plants (n=3 for each treatment) were also used to study stomatal properties (as in the subsections 4.4.1 and 4.4.2). In this paired experiment, the correlation analysis in Table 4.2 reveals that many of the stomatal-epidermal parameters still show a significant relationship, similar to that in Table 4.1 that used a larger sample size (n=11).
When taken overall, the correlation matrix in Table 4.2 shows no obvious relationship between stomatal properties with assimilation and gs values. However, a careful look at Fig. 4.1 shows that although SCA size (Fig. 4.1A) of P5 transferred leaves had similar size to LL grown leaves, the SD (Fig. 4.1I) of the three measured plants did have higher mean density among all treatments. This is consistent with Franks and Beerling (2009) who suggested (based on fossil samples) that high SD could play a role in achieving high gsmax. Correlation analysis between stomatal properties and iWUE showed that guard cells width (GCW) and stomatal complex width (SCW) had a significantly positive relationship with iWUE400 but SD showed almost no correlation (Table 4.2). Once again the data pooled for GCW and SCW showed significant correlation with iWUE400 but not stomatal density (Table 4.2).
Table 4.2:
Pearson’s correlation coefficients (r) among the pooled values of stomatal-epidermal against leaf physiology properties (n=3 for each parameter). Single asterisk indicates correlations which are significant at p<0.05 confidence limit while double asterisks * indicate correlations which are significant at p<0.01 confidence limit.
Stomatal-Epidermal | Physiology | |||||||||||||||||
Stomatal-Epidermal | SCA | SD | GCW | SAL | SCL | SCW | CFW | SPA | %SF | A400 | Amax | gs400 | gsmax | iWUE400 | iWUEmax | |||
SCA |
|
|
|
|
|
|
|
|
|
|
|
|
|
| ||||
SD | -0.64** |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |||
GCW | 0.27 | 0.07 |
|
|
|
|
|
|
|
|
|
|
|
| ||||
SAL | 0.74** | -0.58** | 0.09 |
|
|
|
|
|
|
|
|
|
|
| ||||
SCL | 0.80** | -0.66** | 0.15 | 0.91** |
|
|
|
|
|
|
|
|
|
|
| |||
SCW | 0.57** | -0.05 | 0.49* | 0.29 | 0.17 |
|
|
|
|
|
|
|
|
| ||||
CFW | 0.28 | -0.07 | 0.39 | 0.17 | 0.06 | 0.53** |
|
|
|
|
|
|
|
|
| |||
SPA | 0.65** | -0.43* | 0.31 | 0.70** | 0.65** | 0.28 | 0.37 |
|
|
|
|
|
|
| ||||
%SF | -0.37 | 0.73** | 0.20 | -0.21 | -0.39 | 0.20 | 0.32 | -0.15 |
|
|
|
|
|
| ||||
Physiology | A400 | 0.17 | 0.16 | 0.00 | 0.19 | 0.13 | 0.35 | 0.03 | -0.17 | 0.22 |
|
|
|
|
|
| ||
Amax | 0.18 | 0.32 | 0.21 | -0.03 | 0.04 | 0.39 | 0.05 | -0.12 | 0.23 | 0.73** |
|
|
|
| ||||
gs400 | -0.08 | 0.02 | -0.34 | 0.19 | 0.06 | -0.12 | -0.10 | -0.19 | 0.09 | 0.70** | 0.10 |
|
|
|
| |||
gsmax | -0.05 | 0.01 | -0.31 | 0.19 | 0.06 | -0.04 | -0.01 | -0.20 | 0.11 | 0.72** | 0.12 | 0.98** |
|
| ||||
iWUE400 | 0.34 | 0.12 | 0.59** | 0.15 | 0.21 | 0.61** | 0.35 | 0.32 | 0.31 | 0.09 | 0.47* | -0.51* | -0.44* |
| ||||
iWUEmax | 0.14 | 0.07 | 0.34 | 0.04 | 0.11 | 0.24 | 0.27 | 0.21 | 0.28 | -0.11 | 0.32 | -0.64** | -0.60** | 0.73** | ||||
The question is: why did GCW not correlate with any other gas exchange parameter? Since iWUE gives information about both assimilation and conductance, it seems that GCW influences the CO2 influx to fulfil carboxylation demand while at the balancing water transpired to optimise the leaf temperature.
Since GCW showed significant correlation with a number of leaf physiology parameters, a contour plot from the pooled data was prepared to see how stomatal size (GCW) and patterning (SD) might predict iWUE at ambient CO2 level. The plot (Fig 4.13) suggests that a combination of GCW >4.6 µm and SD within 250-350 mm-2 would produce 70-120 µmol CO2 mol H2O-1 iWUE400. This was the usual range observed in leaves grown in continuous HL conditions (Fig. 4.10A and 4.10C).
Figure 4.13
Contour plot of intrinsic water use efficiency at ambient CO2 level (iWUE400) as a function of guard cells width (GCW) and stomatal density (SD). Note that individual data point for high light (HL) grown leaves are clearly marked. The unit for the colour scale on the top right is µmol CO2 mol H2O-1.
4.5 Discussion
The results described above indicated relatively high variation in both stomatal and cell file properties on transfer at P1 and P3 stages relative to later stages of leaf development, indicating that stomatal phenotypes are highly plastic with respect to the prevailing environmental conditions during these developmental stages. The ability of plants to acclimate to the current environment, especially irradiance, by undergoing morphological and physiological changes, has been well documented in dicots, such as A. thaliana (Kouřil et al., 2012), and monocots, such as Festuca arundinacea (Wherley et al., 2005). In rice, since the P1 and P3 stages are concealed within layers of outer leaf sheaths, the current light conditions signals are presumably sensed by older leaves and relayed to the actively developing primordia (known as remote long-distance signalling control, Thomas et al., 2004). A number of possible signals (peptides, phytohormones, redoxes and sugars) have been postulated as the ‘batons’ in the signal relay from mature leaves (Coupe et al., 2006; Yoshida et al., 2011) which trigger response is in the developing leaf, but the identity of the signal(s) remains elusive. With respect to the responding leaf, our data indicate that the leaf primordium surface up to P3 stage only comprises undifferentiated epidermal cells, suggesting that shifts in the environment during the 2-4 day developmental window prior to overt differentiation during P4 stage can have major effects on final leaf stomatal characters. Our new observations support and extend previous observations that relatively early stages in leaf development are involved in the acclimation processes which set the limits of potential performance of the mature leaf (Oguchi et al., 2003; Murchie et al., 2005).
The data reported here show that various stomatal size parameters are, in general, larger for HL grown leaves than LL grown leaves, in agreement with Hubbart et al. (2012) who also reported significantly bigger (longer) stomata on the abaxial surface in rice grown under HL condition. Since the stomatal complex in rice consists of four cells (two subsidiary cells flanking two guard cells), its area (SCA) is strongly determined by other dimensional parameters such as stomatal complex length (SCL) and stomatal complex width (SCW). Structures in the complex that play an important role in determining gas influx, such as guard cell width (GCW) and stomatal pore area, are associated with parameters of stomatal complex size, so that when all of these stomatal size-related parameters analysed, they show a significantly correlation (Table 1). SCA and SCW must to a certain extent be set by cell file width and it is noticeable that HL leaves have wider cell files than LL leaves. Thus this parameter (cell file width), set at the level of the whole leaf, will have a significant impact on the width parameter of all stomata formed within the leaf.
After transfer from HL to LL, P1 and P3 stages of rice leaf development define a sensitivity window for various aspects of photosynthesis, (Van Campen et al., 2016), and during this window the plant is able to alter stomatal size parameters in order to adjust to the prevailing LL condition. It is noticeable that stomatal size measurements made in leaves after P1 and P3 transfer consistently showed significantly higher variances compared to later P5 transferred leaves. This fits with the idea that P1-P3 stage leaves are highly plastic with respect to response to altered environment, possibly leading to over- and under-shoot in stomatal size parameters. By the P5 stage this plasticity appears to be greatly decreased, leading to a more limited response (with accompanying lower variance in the data measured). The reverse LL to HL transfer appeared to be less sensitive with respect to altered stomatal size parameters, although it is worth noting for this direction of transfer all plastochron-stages showed a significant change in GCW.
Together with size, patterning of stomata is frequently coupled to the determination of gas exchange capacity for a leaf. Similar to findings reported by Hubbart et al. (2012) I found that although stomatal density (SD) for HL grown leaves tended to be slightly higher than those of LL grown leaves, the difference was statistically not significant. There was a large spread of data points in the HL to LL transfer experiments, especially in P1 and P3 transfers. These data suggest that rice can respond to a change in light environment during P1 and P3 stages but this leads to a general increase in noise in the cell division process rather than directed change in overall stomatal density. As mentioned above, the number of stomata per area will be determined to some extent by the number and size of cell files within which stomata can form. Any variation in these parameters will be to some extent countered by the normal functioning of the stomatal patterning system (leading to a normal average stomatal density) but the greater the variation in the background field of cells in which patterning occurs will, as a consequence, tend to lead to a greater variation in the densities observed.
Interestingly rice appears to show a highly consistent number of cell files that contain stomata within any single interveinal gap (maximum of four per any interveinal gap observed) and this value is independent of the actual number of cell files present in the interveinal gap. This characteristic, combined with the leaf’s ability to change CFN and CFW leads to a significantly higher stomatal file percentage (%SF, Fig. 4.1G) for HL grown leaves compared to LL grown leaves. In fact this %SF parameter is another way to measure stomatal patterning (in addition to SD). It follows the trend observed for other parameters in that transfer at P3 stage appears to be the cut-off point at which alteration in final parameter observed in mature leaves can occur. Again the leaves transferred from LL to HL transfer seem to be less responsive.
The physical changes observed in leaf primordia following change in light environment are likely to be attributable to changes in gene expression. For transfer from HL to LL condition, Narawatthana (2013) reported downregulation of certain genes involved in cell and tissue morphogenesis, such as one of the of STRUBBELIG-RECEPTOR FAMILY (SUB) members (SRF8) and SCARECROW (SCR). Even though SUB is thought to be important in the control of cell division orientation plane, as well as controlling cell number shape and size (Chevalier et al., 2005), no confirmed literature could be found for SR8 function except that it has been implicated in peroxidase activity and sterol biosynthesis (Eyüboglu et al., 2007). However there are numerous reports in the literature on the importance of SCR in many aspects of cell development, including the regulation of cell proliferation and guard cell expansion (Dhondt et al., 2010), Kranz’s anatomy establishment and asymmetric division in stomatal development (Kamiya et al., 2003). Dhondt et al. (2012) showed in their work that the scr mutant had reduced cell division rate and guard cell area. Thus the changes observed in stomatal characteristics after P3 transfer could be related to altered SCR expression levels. In this analysis it was striking to see the similarity in SD between HL and LL grown leaves. This finding is in line with other findings in rice which also found no SD difference under HL and LL systems (Hubbart et al., 2013) but contrasts with the finding in Arabidopsis (Coupe et al., 2006). These differences may be related to differences in the control of stomatal formation in monocot and dicot systems.
In the second section of this study I performed paired transfer experiments in order to investigate the relationship between stomatal phenotypes and different aspects of leaf physiology, particularly assimilation rate (A), stomatal conductance (gs) and intrinsic water use efficiency (iWUE). In the HL to LL transfer, photosynthetic capacity measured at 400ppm CO2 (A400) increased by 25% for HL grown leaves compared to LL grown leaves. A similar trend was also observed for maximum assimilation (Amax) with an increase of 68.6%. HL grown leaves in general produce sun leaves with characteristics such as higher Rubisco and chlorophyll a/b ratio contents (Hubbart et al., 2013), lower levels of light harvesting chlorophyll a/b proteins as well as less thylakoids stacking compared to LL grown leaves (Lichtenthaler et al., 1981; Lichtenthaler et al., 2007).
One important distinction between physical acclimation (e.g., stomatal size) and biochemical acclimation (e.g., carbon assimilation) is that even after relatively late P5-stage transfer the leaves were still able to modulate A400 and Amax to match leaves grown in continuously LL conditions. P5 stage is marked by the first appearance of the leaf blade from the encircling leaf sheath layers. This point onwards to the day of measurement (mature leaf, P6 stage) usually takes about 5-6 days. This relatively long period of time had minimal impact on stomatal-epidermal phenotypes but was sufficient to turn the transferred leaf into ‘shade-leaf’ physiologically thus resulting in A-Ci curves similar to those observed in LL leaves (Fig. 4.4iv). Transfer following full leaf expansion does not change Rubisco content (Murchie et al. 2005), but since P-transferred leaves were still extending, this did change the Rubisco content as shown by reduced Vcmax values than HL leaves (Fig. 4.5C) although it was still higher compared to P1 and P5 transferred leaves. Thus the relatively higher assimilation rate by P5-transferred leaves, almost as high a HL leaves, can be explained by a correspondingly lower relative stomatal limitation (Fig. 4.5D) which supports CO2 diffusion, thus relatively more is available for Rubisco carboxylation. In the reverse LL to HL transfer experiment similar mean A-Ci curves (Fig. 4.5iv) were recorded for each plastochron stage transfer which practically matched the mean A- Ci curve for HL grown leaves. Analysis of A400 and Amax also supports this by showing that all transfer treatments led to values that were statistically similar to the leaves grown continuously in HL. For this transfer direction, the point of P5-stage leaf emergence to the day of measurement usually takes about 3-4 days. Despite this relatively shorter acclimatization period compared to the H-L transfer discussed earlier, the resulting leaves still have a final photosynthetic ability comparable to HL grown leaves. The enhancement of assimilation can be caused to a certain extent by parallel changes in some stomatal parameters, such as higher SD and larger SCW which show a positive correlation with Amax (r=0.32 and r=0.39 respectively). More importantly it has been shown that even mature rice leaves have the ability to alter Rubisco and chlorophyll a/b ratio contents following transfer from LL to HL environment (Murchie et al., 2005) thus rendering a high photosynthetic rate.
In addition to biochemical processes, rice leaves can also display changes in gross leaf morphology (thickness) and constituent cell size which might also influence photosynthetic performance. For example, previous work has shown that after exposure to high irradiance rice leaves are thicker (Murchie et al., 2005). Work from our own group (Van Campen et al., 2016) has shown that after transfer from LL to HL at P1 and P3 stages mature rice leaves have a similar thickness to leaves from plants maintained continually under HL. At later stages (P5) leaf thickness does not change appreciably after change in irradiance. Therefore for P5-transferred leaves change in mesophyll thickness cannot be the reason for the comparable Amax measured, suggesting a larger role for biochemical changes, potentially linked with altered stomatal performance (rather than altered stomatal characteristics of size or density).
The analysis of gs-Ci curves also provides some interesting findings on how stomatal structures affect gs. In doing this the curves are divided into two phases: i) stomatal opening and ii) stomatal closing. In the first phase of the curve, HL grown leaves generally have a steeper slope than LL grown leaves, indicating rapid stomatal opening to increase CO2 influx for active carbon assimilation activity. In all curves the entry into the second phase, that is stomatal closing, occurred before ambient CO2 level. At this point assimilation is still in the first phase of A-Ci curve (Rubisco limited) so, theoretically, the leaf should just allow the stomata to open without any influence of water conservation (since the plants were grown hydroponically). This tight stomatal aperture regulation even in the presence of an abundance of water indicates a built-in program to maintain water use efficiency (WUE) under any water status condition. The decrease in gs values in higher CO2 level is normal in C3 and C4 plants (Osborne and Sack, 2012). Unlike the A-Ci curves pattern described in the previous section, which seems not in line with the physical measurements of stomata in the transfer experiments, gs-Ci curves have the tendency to follow the expected trend where P1 and P3 transfer stages set the developmental window for the plant to alter and match the gs value normally obtained in the plants grown continuously under the same light condition. However (and unexpectedly) correlation analysis failed to show any potential relationship between any stomatal phenotypes and gs values (Table 4.2). The lack of correlation is probably due to the saturating high light conditions used during measurements, which were different from those in which the plants were grown.
It is worth to relate the P-transferred stages with the corresponding morphological development that occurs. P1, the earliest form of the developing primordium (Table 1.1,) comprises undifferentiated founder cells and is the most responsive to the altered light condition. P3 stage is marked by the initiation of epidermal specific cells, some of which will eventually form stomatal cell rows, as outlined by Liu et al. (2009, Fig. 1.8). Hence P3 is a critical stage during which stomatal properties such as size and density are set. Later transfer at P5 stage will not do much in altering the physical properties of stomata, thus any acclimation by the leaf at this stage (or later) will come primarily from biochemical/physiological properties. One important finding worth highlighting is that rice stomata in LL grown leaves (which are relatively small) respond minimally to changing CO2 level, thus they do not follow the general rule that small stomata show a faster response than larger stomata in terms of opening and closing (Hetherington and Woodward, 2003; Franks and Beerling, 2009; Lawson and Blatt, 2014). This finding warrants further investigation.
The final aspect of leaf physiology assessed in this chapter deals with iWUE, which can be regarded as an indicator of overall leaf performance. This study confirms that iWUE generally increases with high CO2 reflecting a relatively higher A and lower gs (Tricker et al., 2005; Ainsworth and Rogers, 2007). Despite the mixed patterns observed in some of the transfer experiments, rice leaves in general seem to be very good at compensating, thus maintaining general performance. Thus, irrespective of whether the transfer takes place at the earliest (P1) or latest (P5) stage, the mature leaves still manage to adjust iWUE so that it is adapted to the final light condition. This indicates that rice has a good ability to maintain a good synchrony between A (biochemical) and gs (structural) properties to bring about adjustment to the prevailing environment. It is interesting to observe that iWUE can be predicted essentially by one stomatal character, namely guard cell (GC) width. This is somewhat expected as previous studies have confirmed the importance of GC size in controlling stomatal movement and its positive association with photosynthesis and WUE (Xu and Zhou. 2008; Lawson and Blatt, 2014). Nevertheless, the knowledge from this study might be used to target particular parameters to generate leaves with good iWUE. Based on the predictions indicated by the contour plot using GCW and SD (Fig. 4.11) for the best iWUE at ambient CO2 level, it seems to be always beneficial to have large guard cells but stomatal frequency must also be moderately high if rice leaves are to have a high iWUE.
The control of stomatal properties in rice (Oryza sativa L.) and their influence on photosynthetic performance
June 2017
DOI:
Thesis for: Doctor of Philosophy
Advisor: Prof. Andrew J. Fleming
Attribution 4.0 International — CC BY 4.0 - Creative Commons
Commenti