Chapter 5

INTRODUCTION

Chronic ozone effects on plants are caused by long term exposure, over several weeks, months or seasons, to frequent, relatively low background ozone concentrations. In the UK these background concentrations are generally between 10 and 30 nmol mol-1 (PORG, 1993). The effects of chronic ozone exposure over a whole growing season are manifest as changes in plant growth, productivity and quality (Manning and Krupa, 1992). Chronic ozone effects can be detected as a reduction in net photosynthesis and stomatal conductance, similar to the effects of acute ozone exposure, but smaller in magnitude (Sanders et al., 1992). Phytotoxic ozone concentrations are only marginally higher than present background concentrations, and as background ozone concentrations are increasing at rates which may exceed the rates of adaptation by genetic and phenotypic plasticity within many plant populations, there is a need to assess the effects of expected future concentrations of background ozone on plant growth and crop productivity (PORG, 1993). An estimated 10 to 35% of the world's grain production occurs in areas where ozone may reduce crop yield (Chameides et al., 1994). Wheat accounts for approximately 41% of the total cereal production in Britain, and significant reductions may be expected in wheat yields during warmer growing seasons, associated with higher ozone concentrations (Brown et al., 1995). Furthermore, the rise in tropospheric ozone concentrations may be expected to continue to increase for the foreseeable future, as the precursors of ozone, hydrocarbons (both anthropogenic and organic) and nitrogen oxides are not under direct control. Therefore, it is necessary to be able to predict, not only the effects of an increasing frequency of acute ozone episodes on plant photosynthesis, but also the effects of increased background concentrations of ozone on plant productivity.

Rising ozone concentrations are not the only changing atmospheric conditions that future crop growth will encounter. Carbon dioxide concentrations are also increasing, as are temperatures (Schimel et al., 1996). Therefore it will be necessary to be able to evaluate the interactive effects of increasing ozone concentrations and elevated [CO2] on photosynthesis and crop production in order to predict and plan for crop yield under future climate conditions (Long, 1991; 1994). Under elevated [CO2] stomata close to maintain a constant ratio of Ca:Ci, found to be about 0.7 for C3 species under a leaf to air water pressure deficit of <2 kPa (Wong, 1979; Long, 1994). Bearing this in mind, the expected interaction between the effects of elevated carbon dioxide concentrations and ozone exposure would be a reduction in the uptake of ozone into the leaf, due to stomatal closure resulting from elevated [CO2]. So, a reduction in the damaging effects of ozone would be expected under conditions of elevated [CO2]. This protective effect of CO2 has been found effective for chronic ozone exposure in wheat (McKee et al., 1995; Fangmeier et al., 1996). An additional protective mechanism in wheat against chronic ozone exposure under elevated [CO2] has also been suggested, due to a shift in the process limiting photosynthesis under elevated [CO2] (McKee et al., 1995). Under ambient [CO2] the process of photosynthesis occurs at the point of inflexion on the A/Ci curve, where the limiting factors of Rubisco and RubP regeneration limitation are least limiting, and thus photoassimilation is conducted with maximum efficiency. Under elevated carbon dioxide concentrations, however, control is removed from Rubisco to RubP regeneration. Thus, under moderately elevated concentrations of ozone, elevated [CO2] increases the efficiency of carboxylation. For leaves with just sufficient Rubisco to support the maximum rate of photosynthesis at current ambient [CO2] Rubisco would be in excess at elevated [CO2] (Woodrow, 1994). Therefore, assuming acclimation to elevated [CO2] does not result in a return to Rubisco levels that are just sufficient for carboxylation, at elevated [CO2] leaves could tolerate some loss of Rubisco due to ozone, without an effect on rates of photosynthesis (McKee et al., 1995).

Prediction of the interactive effects of increasing concentrations of ozone and CO2 on crop productivity may be attempted through a process-based model to simulate the effects of chronic ozone exposure on photosynthesis, in order to provide a means of predicting chronic ozone effects under varying environmental conditions, including elevated CO2, which can subsequently be combined with the biochemical mechanistic model of photosynthesis of Farquhar et al. (1980). It has been shown in the previous Chapter that a simple mechanistic model can be used to predict the effects of acute ozone exposure on photosynthetic rates. Can chronic effects be similarly modelled by this simple approach? If so, the combination of such a model with the biochemical mechanistic models of CO2 assimilation of Farquhar et al. (1980) and the phenomenological model of stomatal conductance of Ball et al. (1987) would provide a strong hierarchical model to simulate the interactive effects of elevated concentrations of O3 and CO2 on photosynthesis, as a first step in predicting the effects of future climates on crop productivity (Figure 4.9, p.121). Testing the model against data at [CO2] of 700 µmol mol-1 measured by I.F. McKee may also determine whether protection against ozone damage, in addition to that provided by stomatal closure, was afforded by the shift in factors limiting photosynthesis at elevated [CO2], at the fifth leaf stage.

The over all aim of this chapter is to present a model to simulate the effects of chronic ozone exposure on wheat, and to use this model to predict the interactive effects of elevated background ozone and carbon dioxide concentrations on wheat leaf photosynthesis and wheat productivity. The model is based on the hypothesis that the damage caused to wheat photosynthesis by chronic ozone dose can be related to an effective ozone dose, above a threshold flux.

There are five objectives to this chapter. First, to test the hypothesis that the damage caused to wheat photosynthesis by chronic ozone dose can be related to an effective ozone dose, above a threshold flux, using the unpublished data of I.F. McKee. Second, to construct a model to simulate the effects of chronic ozone exposure on wheat photosynthesis. Third, to use the model to simulate the interactive effects of ozone and carbon dioxide on leaf photosynthesis in wheat. Fourth, to test the model against data collected by I.F. McKee (unpublished) for the interactive effects of elevated [CO2] and chronic ozone exposure. Fifth, to use a simple canopy model to scale up the leaf level model to the canopy level, in conjunction with an empirical relationship between leaf area index and mean daily temperature to drive a wheat growth model, and so predict the effects of chronic ozone exposure and elevated [CO2] on wheat productivity.

MODEL THEORY AND DEVELOPMENT

Although the magnitude and biochemical mechanism of ozone damage to plants may differ for acute and chronic exposure, the route of entry of ozone into a plant leaf is common to both types of exposure, namely, via the stomata. Therefore, it is reasonable to assume that the model mechanism presented in the last chapter, used to simulate the effects of acute ozone exposure on wheat photosynthesis, might also apply to chronic ozone exposure. That is, damage is only caused once the flux of ozone into the leaf exceeds a critical rate of delivery (Heath, 1994). Consequently, a linear relationship between an effective ozone dose above a threshold flux (FO3(0)), and the initial site of damage to photosynthesis, Vcmax, might be expected for chronic exposure effects. This expectation is supported by the observation that the decrease in stomatal conductance observed in plants grown with elevated [O3] can be eliminated if the intercellular CO2 concentration is decreased to the level observed in control plants. This result suggests that decreased stomatal conductance in elevated ozone concentrations can be accounted for by the increase in Ci that results from decreased photosynthetic capacity, without the need to invoke any direct effect of ozone concentration on the stomata (McKee et al., 1995). Hence, the process model of ozone damage for acute ozone damage (Figure 4.1, p.109) may also be applicable to chronic ozone damage.

Data collected by I.F. McKee (unpublished) on spring wheat (Triticum aestivum L. cv. Wembly) were used to test whether chronic ozone effects are dependent upon the accumulated dose above a critical rate of ozone delivery into the leaf. The plants used to obtain the data were grown and exposed to [O3] and [CO2] in fully controlled environments as described by McKee et al. (1995). Spring wheat was exposed to four levels of ozone concentration and two levels of CO2 concentration within controlled environment growth cabinets. The data consisted of the effects of ozone on in vivo maximum rate of Rubisco carboxylation and stomatal conductance of the fifth leaf of wheat exposed to diurnal, sinusoidal cycles of ozone concentration for 16 days, at CO2 concentrations of 350 µmol mol-1 and 700 µmol mol-1. The model would be based on the data collected at ambient [CO2] only. The data collected at elevated [CO2] would be used later for testing the model's predictions of the interactive effects of increased concentrations of ozone and CO2. The concentrations of ozone were controlled in a sine wave form during the photoperiod, peaking at <5, 30, 60 or 90 nmol mol-1. The data at ambient [CO2] was used to determine a threshold flux (FO3(0)) and to estimate a coefficient of ozone damage (Kz).

As stomatal conductance was only measured at the beginning and end of this 16 day period, the mean of the control and the final values of stomatal conductance were used in calculations to estimate the total ozone dose for each [O3]. As the actual pattern of decline in stomatal conductance with time could not be incorporated into the calculations, the use of the mean stomatal conductance was expected to give an overestimation of total dose and effect on Vcmax, particularly at higher doses. Providing the effective ozone dose hypothesis holds, a linear relationship between the predicted decline in Vcmax and measured decline in Vcmax would be expected, but with a gradient greater than unity. Therefore, the ozone coefficient, determined by the slope of the relationship between reduction in Vcmax and effective ozone dose, would need to be adjusted.

Using a simple canopy model, the effects of chronic ozone exposure at the leaf level can be scaled up to the canopy level, and thus indicate effects of increasing ozone and carbon dioxide concentrations on a wheat canopy. An extension of the canopy model to accumulate the effects of chronic ozone exposure over the time period of a growing season would allow simulations of the interactive effects of [O3] and [CO2] on wheat productivity. The growth of wheat through a growing season can be simply simulated as the accumulated assimilation rate less growth and maintenance respiration. The simulated change in leaf area index, based on the parameters of Porter (1984), was related to thermal time, calculated from the mean daily air temperature.

5.3.1 Determination of threshold flux and coefficient of ozone damage

The values of FO3(0) and Kz at ambient carbon dioxide concentrations were determined using the data measured by I.F. McKee (unpublished). The data consisted of final values of Vcmax and stomatal conductance, after 16 days exposure to four levels of ozone concentration. Due to the absence of data of the variation of stomatal conductance during the experiment, mean values of stomatal conductance were estimated. The total accumulated dose of ozone entering the leaf was calculated at hourly intervals, as the product of ozone concentration and mean stomatal conductance to ozone, accumulated over the period of 16 hours for 16 days, as described by Equation 5.1, Table 5.1 (Figure 5.1).

The sine wave variation of ambient ozone concentration is described by Equation 5.2, Table 5.1, p.146, where peak ozone values were <5, 30, 60 and 90 nmol mol-1. Where peak ozone concentrations were <5 nmol mol-1 the peak [O3] was assumed to be equal to 3 for the purpose of O3 dose calculations. Stomatal conductance to ozone (gz) was calculated from the stomatal conductance to water using the conversion method of Laisk et al. (1989), Equation 5.3, and the mean gz was calculated as the mean of the control stomatal conductance at [O3] of <5 nmol mol-1 and the final stomatal conductance, after 16 days exposure to ozone. The measured Vcmax value for 350 µmol mol-1 CO2 concentration, was 35.7 µmol m-2 s-1 (measured by I.F. McKee). By setting the value of Vcmax to 35.7 µmol m-2 s-1, and the coefficient for stomatal conductance g(1), to a value of 28.9, the measured control value of stomatal conductance of 561 mmol m-2 s-1 could be simulated, from the equations of Table 5.2.

Within the model, the stomatal conductance is solved iteratively, by the same method as that used by the acute ozone model (Figure 4.9, p.121). First the rate of CO2 assimilation is calculated and the values of Ci and gs are determined iteratively. The resulting stomatal conductance values are used to calculate the ozone uptake, and the decline in Vcmax is predicted from the chronic ozone model equations.

Table 5.1: Model equations to simulate the effects of chronic ozone exposure on wheat photosynthesis and productivity.

Table 5.2: Table of parameters used for leaf ozone model simulations.

This decline in Vcmax is used by the photosynthesis model equations to calculate the stomatal closure, but because this is also dependent upon Ci, both gs and Ci are again solved iteratively. Another iteration between the gs, ozone uptake and effect on Vcmax is performed before a final value for stomatal conductance at the end of a time interval can be determined.

A linear relationship was sought between effective ozone dose and reduction in Vcmax, by calculating effective ozone doses using different values of FO3(0), and plotting the effective ozone doses for the four ozone regimes against the measured reduction in Vcmax, Equation 5.4. The various values of FO3(0) deducted from the instantaneous flux, during the calculation of accumulated effective ozone dose, were between 4.5 and 7.5 nmol m-2 s-1. The effective ozone dose is calculated as the accumulated dose above a threshold flux of ozone entering the leaf (Equation 5.5).

The value of FO3(0) that gave the maximum correlation between the percent reduction in Vcmax and effective ozone dose, was found to be 6 nmol m-2 s-1, at ambient CO2 concentrations (Figure 5.2). The slope of the relationship between reduction in Vcmax and accumulated effective ozone dose (measured in units of µmol m-2 h-1) gave an estimated value of the Kz to use in the relationship of Equation 5.4, to be inter-linked with the macroclimate and leaf photosynthesis model equations previously described in Chapter Two, p.77 and p.42, and listed in Appendix A, pp.196 and 197 (Figure 5.3). The estimated Kz, or slope of reduction in Vcmax/F'O3eff was adjusted to give a closer fit between predicted and measured data to the 1:1 slope (Figure 5.4). Resulting parameter values are listed in Table 5.2. Figure 4.9, p.121 shows how the ozone models fit into the structure of the whole model.

5.3.2.1 Simulated interactive effects of ozone and elevated [CO2]

To test the model, the simulations of the interactive effects of ozone exposure and elevated carbon dioxide concentrations were compared with the data measured at elevated [CO2], not previously used in model construction (McKee, unpublished). To simulate the interactive effects of ozone exposure and elevated carbon dioxide concentrations, the model was parameterised with the ozone exposure regimes described by Equation 5.3, at peak values of 3, 30, 60 and 90 nmol mol-1, together with [CO2] of 700 µmol mol-1. The independent measured value for Vcmax was 32 µmol m-2 s-1 for 700 µmol mol-1 [CO2], and this value was subsequently parameterised into the model for model testing. The stomatal conductance coefficients, g(0) and g(1), for 700 µmol mol-1 [CO2] were kept the same as those for ambient [CO2], 81.1 and 28.9, respectively. This gave the predicted control value of stomatal conductance at 700 µmol mol-1 [CO2] of 238 mmol m-2 s-1, a value within 5% of the measured value, 250 mmol m-2 s-1. Predicted percentage reductions in Vcmax were compared with measured percentage reductions, using FO3(0) equal to 6 nmol m-2 s-1.

5.3.2.2 Testing the model against cv. Avalon

The model was tested against A, gs and Vcmax determined after 7 days exposure to ozone, using data collected by Farage (unpublished) for chronic exposure to winter wheat (Triticum aestivum L. cv. Avalon). The ozone exposure regime was different to that used by I.F. McKee et al. (1995), as were the experimental procedures, equipment, and cultivar of wheat used, and would therefore provide a valid test for the applicability of the model to wheat in general. The [O3] regime consisted of 80 nmol mol-1 of ozone for 7 hours a day, in a square wave pattern, for 18 days from the sowing date. The effects of ozone exposure on the second leaf, in terms of reduction in Vcmax, was measured after 18 days. The second leaf was assumed to have been exposed to ozone for 10 days, allowing 8 days for germination and leaf initiation. The initial value of Vcmax was 65.09 µmol m-2 s-1, but unfortunately, the stomatal conductance was not measured in situ. Due to the lack of available data containing the required measurements of both stomatal conductance and values of Vcmax, this data point will be used as a preliminary test of the model.

As the stomatal conductance was not measured in situ, the model would be tested in two ways. Test One: using the Vcmax value measured by P.K. Farage for the cultivar Avalon (65.09 µmol m-2 s-1) and stomatal conductance coefficients observed for the data of I.F. McKee on the cultivar Wembly. Test Two: using the values of Vcmax (35.7 µmol m-2 s-1) and stomatal conductance coefficients observed for the data of I.F. McKee, as a test of the applicability of the model to the effects of chronic ozone on wheat photosynthesis to a different cultivar (Table 5.2). The threshold flux and coefficient of ozone damage was assumed to be those that applied to the data of I.F. McKee in both model simulations. Thus, g(0) and g(1) were 81.1 and 28.9, respectively, the threshold flux was 6 nmol m-2 s-1 and the coefficient of ozone damage was 0.009 (Table 5.2).

5.3.3.1 Parameterisation of ozone variation during the growing season

For the purpose of model predictions the pattern of variation in background ozone concentrations during a growing season were based on the data of ozone concentrations at Sibton, during 1992 (DOE, 1994). Sibton is described as open, flat cereal farmland and therefore was chosen as a suitable rural site on which to base model simulations. A simple sinusoidal relationship was fitted to the monthly mean ozone concentrations of 1992, Equations 5.6-5.9, Figure 5.5. A further sinusoidal relationship was fitted to the mean hourly ozone concentration for Sibton at hourly intervals during a summer's day Equation 5.9, Figure 5.6 (DOE, 1994).

The values of parameters for ozone concentration calculations are listed in Table 5.3.

5.3.3.2 A simple canopy model to simulate wheat growth

The model given in Table 5.1, Equation 5.10, and parameters in Table 5.2 were used to relate wheat development to accumulated thermal time, above a base temperature. A relationship between the change in leaf area index (lcan) of the late sown wheat cultivar Hustler for three consecutive years (Porter, 1984), and the mean daily air temperature of the UK (from Chapter Three, p.80) was derived, using the accumulated thermal time above a base temperature of 4oC (Porter, 1984), scaled down by a factor of 12 (Figure 5.7), Equation 5.10. In this way, the change in leaf area index up to anthesis was simulated by the new model presented here, from the accumulated thermal time calculated from the simulated change in mean daily air temperature for the UK. The leaf area index was assumed to decline linearly after anthesis, arbitrarily set at the Julian Day of 178.

The new model was first used to simulate the above ground biomass, estimated as the product of the proportion of above ground biomass to total biomass (0.7) and total weight (WT) (Equation 5.11). The total weight is the calculated as the CO2 assimilation accumulated over the growing season, less maintenance and growth respiration, multiplied by the factor to convert the weight of CO2 to carbohydrate weight (0.65) (Equation 5.12). The equations for growth and maintenance respiration are those used in the AFRC wheat model (Porter, 1984), adapted from the method of McCree (1970). Growth respiration, which supports the synthesis of new biomass, is estimated as a proportion of assimilation rate (Equation 5.13). Maintenance respiration supports the turnover and replacement of existing cell components, and is estimated as a proportion of plant dry weight, which also is dependent on mean air temperature (Equation 5.14).

First the above ground biomass of wheat grown was calculated, at ambient [CO2] of 350 µmol mol-1, at both zero [O3] and ozone concentrations simulated for Sibton (DOE, 1994). Then the above ground biomass was predicted for the same ozone regimes, but under elevated [CO2]. The effects of [O3] and [CO2] were simulated as effects on accumulated carbon dioxide assimilation, and feedback effects on LAI were omitted. Parameter values used in the model simulation are given in Table 5.3 and Appendix II.

Table 5.3: Values of model parameters used in wheat growth model simulations.

5.4.1 Interactive effects of increased [O3] and [CO2] on Vcmax

The percentage decline in Vcmax predicted by the model parameterised for the data of I.F. McKee, compares well with the percentage reduction in Vcmax at elevated carbon dioxide concentrations, measured independently by I.F. McKee, Figure 5.8. The value of FO3(0) used was 6 nmol m-2 s-1 and the value of Kz was 0.009, the same values as used for ambient [CO2]. The failure of the model to predict the slight decrease in Vcmax at low ozone doses could have been corrected by the inclusion of an empirical relationship to predict ozone effects at low ozone concentrations. However, as the aim of this study is to produce models that are as mechanistically rich as possible, and due to the relatively small error resulting from its omission, it was decided to exclude the empirical relationship for the effects of ozone at very low doses.

5.4.2 Model testing

The measured reduction in Vcmax was 30.8%. The predicted reduction in Vcmax using the parameter values of Test One (Table 5.2), that is, the control value of Vcmax measured by P.K. Farage (unpublished) was 21.7%. However, the in situ initial stomatal conductance, critical to the calculation of instantaneous ozone flux and an effective ozone dose, was unknown. Using the model parameter values that applied to the data of I.F. McKee, that were used to construct the model to predict chronic ozone effects (Test Two, Table 5.2) the predicted reduction in Vcmax was 25.2%. The prediction of Test Two was closer to the reduction measured by P.K. Farage than the prediction of Test One.

5.4.3 Effects of chronic ozone exposure and elevated [CO2] on wheat productivity

The model predicted an above ground biomass yield of 1.85 kg m-2 under zero ozone concentration, compared to a yield of 1.47 kg m-2 under the simulated chronic ozone concentrations, at ambient [CO2], based on the observed [O3] variation recorded at Sibton in Suffolk, during 1992. This is a decrease of 20.6%, Figure 5.9. Under elevated CO2 the above ground biomass was predicted to be 2.42 kg m-2 for both 0 nmol mol-1 [O3] and the simulated chronic ozone concentrations based on data from Sibton in 1992. This apparently surprising result, that the model predicted no effect of chronic ozone exposure on wheat productivity at elevated [CO2], is less surprising when the low photochemical activity of the year 1992 in taken into account. The year 1992 was not one of the hottest years, and the peak ozone concentrations measured during the summer of 1992 in the UK were substantially lower than during the two heat-wave years of 1989 and 1990 (DOE, 1994).

DISCUSSION

The good fit between measured reduction in Vcmax and effective ozone dose, as illustrated in Figure 5.3, supports the hypothesis that the damaging effects of chronic ozone dose to wheat leaves can be related to a critical rate of ozone delivery into the leaf (Heath, 1994). The relationship between percentage decline in Vcmax and ozone uptake (F'O3tot) is not linear, as the reduction in Vcmax at relatively low dose levels is small (Figure 5.1). Also, the three data points at 350 µmol mol-1 [CO2] above a total ozone dose of 5 mmol m-2, describe a curvilinear relationship similar to that found between ozone uptake and effect on Vcmax under acute ozone exposure (Figure 4.5, p.115). Thus, the first of this chapter's objectives, to test the hypothesis that the damage caused to photosynthesis in wheat by chronic exposure to ozone can be related to an accumulated dose of ozone into the leaf (F'O3eff), above a threshold flux (FO3(0)), has been met. As the hypothesis was consistent with the observations, a process based model could be constructed, allowing the second, third, fourth and fifth objectives to be achieved.

The second objective, to construct a model to simulate the effects of chronic ozone on wheat photosynthesis, was achieved by incorporating the relationship between reduction in Vcmax and F'O3eff into the model equations of photosynthesis and stomatal conductance, Equations of Table A, Appendix I, (Farquhar et al., 1980; Ball et al., 1987; Humphries and Long, 1995). Testing the model against the data of P.K. Farage (unpublished) produced a result that was within 10% of the measured value. However, partly due to the lack of data on in situ stomatal conductance, together with the difficulties of relating results from one experiment to another, further research is needed before the model can be considered to be comprehensively tested.

The model was used to simulate the interactive effects of elevated concentrations of ozone and carbon dioxide on wheat leaf photosynthesis, and subsequent testing of the simulated results against the measured data showed elevated [CO2] to protect against the damaging effects of ozone, by reduced stomatal conductance, thus fulfilling the third and fourth objectives. The fifth objective, to scale up the leaf level model to the canopy level to predict the effects of chronic ozone exposure and elevated [CO2] on wheat productivity produced a result that was, perhaps, unexpected. The model predicted [CO2] elevated to 700 µmol mol-1 to protect wheat above ground biomass productivity from the effects of chronic ozone exposure, such as those that occurred during 1992, at Sibton (DOE, 1994). The result, however, is not so surprising considering the low photochemical activity of the year 1992. The peak O3 concentrations measured during the summer of 1992 in the UK were substantially lower than during the two heat-wave years of 1989 and 1990 (DOE, 1994). In photochemically active years, which are expected to occur with greater frequency in the putative warmer, drier climates of the future, some protection against the damaging effects of chronic ozone exposure at elevated [CO2] might be expected. However, warming has not occurred consistently from year to year, and so, during cooler years in the UK, at the low chronic ozone concentrations used in this study, it is not inconceivable that [CO2] elevated to 700 µmol mol-1 would protect a wheat crop from loss due to chronic ozone exposure, as this model predicts.

In the past, due to the complexity and inter-dependence of ozone dose effects and stomatal conductance, it has been very difficult to define direct relationships between ozone exposure, dose and effect, as evident by the myriad of proposed exposure-response indices, for examples, see Lefohn (1992). Dose-response relationships are recognised to be more relevant, but the dynamic nature of the plant system limits the ease with which such relationships can be identified (Runeckles, 1992). Hence, to date, no mechanistic model of ozone effects on photosynthesis has been constructed. However, crop simulation models have been constructed to predict of the effects of changing environmental conditions on crop yield. For example, Mitchell et al. (1995) tested the AFRCWHEAT1 simulation model against experimental data of the effects of elevated [CO2] and increased temperature on winter wheat. The AFRCWHEAT1 model used however, had a greater empirical content than the model presented here. Mitchell et al. (1995) reported that the inclusion of a more mechanistic sub model of productivity, based on SUCROS87 and model equations of Farquhar et al. (1980), closely simulated the observed crop biomass, thus supporting the aims of this study, the use of models with a high mechanistic content wherever possible.

There are two main limitations of the present model. The first is due to the lack of data for model validation, and thus the uncertainty of the wider use of the model to simulate the effects of chronic ozone exposure on wheat in general. Meanwhile, using the data point measured in a different experiment by P.K. Farage, where the second leaf was exposed to ozone at 80 nmol mol-1 for 7 hours a day, for approximately 10 days, the initial Vcmax was observed to be reduced by 30.8%. Using the parameters from I.F. McKee's study, the model predicted a reduction in Vcmax of 25.2%, which is a promising result. However, further research which includes the measurement of stomatal conductance and Vcmax values during the period of exposure to ozone is required for further model validation, as stomatal closure controls the rate of entry of ozone into the leaf.

The second main limitation, once again, due to lack of data, is the lack of a simulation of recovery time in the model. The present model assumes no overnight recovery of Vcmax and no recovery during successive days of low exposure, although evidence for recovery after exposure has been found (Manning and Krupa, 1992). Model equations to simulate recovery have been proposed and an adjustment to simulate recovery would be easy to incorporate into the model presented here, should future research enable coefficients of recovery to be parameterised for wheat (Runeckles and Chevone, 1992).

The further development of the model to scale-up the ozone effect relationship on wheat photosynthesis at the leaf level to the canopy level and subsequently, to the effects on wheat production, allowed the comparison of ozone and elevated [CO2] effects acting individually and concurrently on wheat yield, in terms of above ground biomass. For the purpose of simulating ozone concentration in this study, the discrepancies between the data of monthly and hourly means of [O3] measured at Sibton in 1992, and the simulated pattern of [O3] variation, were assumed to be of little significance (Figures 5.5 and 5.6). For example, the greatest differences during the year were during the period when leaf area index values were declining after anthesis, and during the day the simulated curve underestimated overnight concentrations, when photosynthesis does not occur (compare Figure 5.5 with Figure 5.7).

The aim of this study was to predict the interactive effects of increasing concentrations of O3 and CO2 concentrations on wheat productivity, in terms of proportionate change of crop production under ambient [CO2] and no ozone. For this purpose it was considered necessary to construct a relatively simple growth model. Thus, the partitioning of biomass into above and below ground biomass components was presumed to be an adequate simulation of partitioning, omitting the simulation of growth stage partitioning used in more complex, empirical wheat growth models, such as the AFRC wheat model (Porter, 1984). The relatively simple partitioning into above and below ground components avoided the ambiguity of the effects of ozone on the shoot to root ratio by the use of an unvarying variable (0.7) to represent the proportion of above ground to total biomass. The ambiguity of effect of ozone on shoot:root biomass arises from the conflicting results reported in the literature. Ozone has often been found to suppress root growth more than shoot growth (Miller, 1987; Woodberry et al., 1994; Barnes et al., 1995). However, this is considered to be an oversimplification, as responses vary with ozone exposure regimes and from population to population (Reiling and Davidson, 1992; Pearson et al., 1995). Even without the inclusion of any feedback effect that chronic ozone might be expected to have on leaf area index, the potential effect of chronic ozone concentrations recorded at Sibton during 1992 on a wheat crop are predicted to be considerable. The model predicts a reduction of 22% from a crop simulated to be grown in air free of ozone, at ambient [CO2] (Figure 5.9). This is very close to the range of crop loss by ozone predicted generally for the UK, by the statistical model of Smith et al. (1995). They predicted a loss of between 5-15% with errors in the range of 2-6%.

The model predicted [CO2] elevated to 700 µmol mol-1 to cause a 30% increase in above ground biomass of a wheat crop, which was predicted to be undiminished by the ozone concentration regime based on those measured at Sibton, in 1992 (Figure 5.9). Further experimental research is required for model validation, and refinement of the model. The model has potential to predict the effects of climate and atmospheric change on wheat productivity.

Future developments of the model should include the incorporation of the effects of ozone during different growth stages of wheat and the effects of climate change on the time scales of wheat growth and development. For example, wheat development time scales can be affected directly by ozone, due to both enhanced senescence and the shortened time periods of developmental stages induced by higher temperatures (Fangmeier et al., 1994; Rozenzweig and Tubiello, 1995; Soja and Soja, 1995). However, the more urgent need before this model can claim to be valid, is for further experimental data to test the model against, and also to begin to develop the model to simulate the small reductions in Vcmax at relatively low ozone concentrations (Figure 5.8).

Although there is a lack of models predicting the effects of ozone on wheat productivity with which to compare the results of this study, a model based on the mass and energy exchange of a forest canopy, using eddy correlation, was used by Amthor et al. (1994), to compare measured and predicted rates of carbon dioxide and ozone uptake by a mixed oak-maple stand. The big-leaf model, which treats the canopy as a single homogenous collection of leaves, under-predicted the amount of ozone taken up by the forest canopy during the morning and midday, which could be accounted for by deposition rates to external surfaces of the canopy. However, the model over-predicted afternoon rates of CO2 uptake and stomatal conductance, which might result from sink feedback (Amthor et al., 1994), but could be due to the accumulated effects of ozone on the photosynthetic apparatus, not accounted for by their model.

[ Home ] [ Up ]

Send mail to humph@essex.ac.uk with questions or comments about this web site.
Copyright © 1998 WIMOVAC Ltd.
Last modified: August 19, 1997