The use of gas exchange characteristics to optimize CA storage and MA packaging of fruits and vegetables

Oxygen consumption as influenced by carbon dioxide
High carbon dioxide (CO ₂ ) concentrations can reduce the oxygen (O ₂ ) consumption rate of a number of fruits and vegetables. This reduction can be modelled by incorporating an inhibition term in an Michaelis-Menten type of model, describing the overall respiration process as a single enzyme reaction. Four types of inhibition can be distinguished: 1. the competitive type, 2. the uncompetitive type, 3. a combination of both previous types and 4. the non-competitive type (after Chang, 1981). Using the inhibition terms a good estimation of O ₂ consumption could be obtained. This supports the use of Michaelis- Menten kinetics for modelling O ₂ consumption.

Depending on the product the statistical analysis gave good results for the competitive and the uncompetitive type of inhibition. Based on gas exchange data only, no distinction between the competitive and uncompetitive type of inhibition could be made. The data suggest the simultaneous existence of both types of inhibition. However, for reasons of simplicity the non-competitive type of inhibition is preferred, giving good results for all the products tested. This non-competitive inhibition term, evaluated in chapter 2, is used in other chapters whenever an influence of CO ₂ on respiration was found.

Oxidative and fermentative carbon dioxide production
Because the main metabolic sources for CO ₂ emission by higher plants are respiration and fermentation, both processes have to be incorporated in a model describing total CO ₂ production. For this purpose one existing model (Peppelenbos et al., 1993) was adjusted and, based on different theories, two models were developed. The adjusted model used O ₂ as an inhibitor of fermentative CO ₂ production, whereas the two new models used the ATP production rate, representing ATP concentration. The difference between the latter two models is that in the first one ATP production is calculated by using only oxidative processes, while in the second one ATP production is calculated by a combination of oxidative and fermentative processes. All models allow for increased CO ₂ production at low O ₂ concentrations, as is often found for several products. The best performance was found for the adjusted model and the new one which used oxidative ATP. The results do not clarify whether increased fermentation rates can be attributed to decreased O ₂ levels or decreased energy fluxes. The approach used, however, enables the calculation of CO ₂ production rates of different types of commodities stored under various gas conditions. This facilitates a better prediction of CO ₂ conditions inside storage rooms and MA packages.

Alcoholic fermentation as influenced by carbon dioxide
Not only respiration can be influenced by high CO ₂ concentrations. In several products this influence is also found on fermentation. This influence was incorporated in the CO ₂ production model based on the inhibition of alcoholic fermentation by O ₂ (the'adjusted' model). Gas exchange rates of mungbean sprouts under various O ₂ and CO ₂ concentrations were used to validate the model. With the modification applied, CO ₂ production rates were described better. Although CO ₂ production at low O ₂ concentrations was reduced by high CO ₂ concentrations, the data showed no influence on ethanol and acetaldehyde levels.

The data obtained indicate large differences between gas exchange rates of different batches of mungbean sprouts. It is suggested that microbial metabolism attributes substantially to total CO ₂ production rates found, and might explain these differences.

The simultaneous measurement of gas exchange and diffusion resistance
A method was developed to measure metabolic gas exchange rates and gas diffusion resistance of apples simultaneously, under various gas conditions. For this purpose the trace gas neon was selected. After closing a flask containing an apple already kept at a specific gas condition, the neon partial pressure was brought to 110 Pa. Changes in oxygen and carbondioxide concentration in the flask were used to calculate gas exchange, and the decrease in neon concentration was used to calculate gas diffusion resistance. The calculated resistance values were compared with data obtained from literature, and estimations of O ₂ and CO ₂ resistance values were made. The method worked well on apples, but this will not necessarily be the case when products are measured with small internal gas volumes.

Functioning of gas exchange models using internal and external concentrations
Based on gas exchange rates and diffusion resistance, internal gas concentrations of apple cultivars Golden Delicious, Elstar and Cox's Orange Pippin were calculated. Internal O ₂ concentrations were 2.3 kPa lower at an external O ₂ concentration of 20.7 kPa for Golden Delicious apples, and about 4.5 kPa lower at 20.1 and 20.4 external O ₂ for Elstar and Cox's apples respectively. Internal CO ₂ concentrations substantially exceeded normal external concentrations of 50 Pa. The Km values found for the three apple cultivars remained significantly different when internal instead of external concentrations were used. This indicates that the apple cultivars measured do not only show biophysical differences (resistance, porosity), but also differences at the biochemical level.

For Golden Delicious apples no difference in model functioning was found when internal or external concentrations were used. In contrast, for Elstar and Cox's Orange Pippin apples the O ₂ uptake and CO ₂ production models showed better results (expressed as R ²) when fitted on external concentrations. It is argued that this might be explained by the experimental setup. For instance the internal O ₂ concentration of Cox's Orange Pippin calculated at the optimal external O ₂ concentration (1.2%) reached 0.01 %. A small change of 0. 1 % in an external O ₂ concentration close to 1 % therefore can change the internal atmosphere from hypoxia to anoxia, which cannot be regarded as an equilibrium situation. The conclusion to be drawn is that also for experimental setups some precalculations using gas exchange rates and diffusion resistances will help to optimize the methods.

Gas exchange characteristics and prediction of optimal gas conditions for CA storage
The applicability of respiratory characteristics to determine optimal O ₂ concentrations for the storage of apples was tested. A comparison was made between gas exchange rates of apples directly after harvest and after a period of storage. Optimal O ₂ concentrations were based on gas exchange data and gas exchange models fitted on the data, using the Anaerobic Compensation Point (ACP) and the Respiratory Quotient Breakpoint (RQB). A third to establish optimal gas concentrations way was comparing total ATP production with estimated maintenance energy requirements, revealing the Maintenance Oxygen Concentration (MOC). ATP production was calculated using gas exchange models. MOC was defined as the oxygen concentration with the minimal ATP production rate necessary for maintaining cell viability. The optimal O ₂ concentrations as established by ACP, RQB and MOC differed considerably. Because ACP values differed from normally advised values, the ACP was unsuitable for a quick determination of the optimal O ₂ concentration of the apples used. The RQB, however, might be suitable, but than the limit used to establish the RQB should be more than 0.5 units higher than the RQ measured in ambient air. The ACP and the RQB were decreased to lower O ₂ concentrations after storage, suggesting that the optimal concentrations decrease during storage. In contrast the MOC was increased after storage, which was in agreement with data as found in the literature. Model calculations indicated the lowest optimal O ₂ concentration for the second (optimal) harvest using the ACP, the RQB and the MOC. It is suggested that research on the relationship between Maintenance Energy Requirements and cell injury will clarify an important part of the changes in optimal O ₂ concentrations (or the tolerance to low O ₂ concentrations) during ageing or maturation of harvested plant tissues.

Fermentation at high oxygen concentrations
Apples were stored at various O ₂ concentrations, ranging between normoxia and anoxia. Gas exchange rates and the production of acetaldehyde and ethanol was measured. A gas exchange model, which distinguishes oxidative from fermentative CO ₂ production, was fitted to the data. The results indicate alcoholic fermentation to be active at all the O ₂ concentrations used, and increasing in importance when O ₂ concentrations are lowered. After calculating the amount of metabolites in the apple tissue from the data measured in air, a close relationship was found between model predictions of alcoholic fermentation rates and measured metabolite production in normoxia and anoxia. In hypoxia, however, the model predicted higher CO ₂ production rates in comparison to the metabolites actually found. Because the model was fitted to CO ₂ production data, this indicates another source or CO ₂ in hypoxia than alcoholic fermentation.

Conclusions
The influence of CO ₂ on O ₂ uptake was investigated, and the known Michaelis-Menten equation given by Chevillotte (1973) was extended with the type of inhibition adequately describing this influence. Models describing fermentative CO ₂ production were developed and combined with oxidative CO ₂ production, enabling the calculation of CO ₂ production of various products under a range of combinations of O ₂ and CO ₂ . Although gas exchange of mungbean and microbial growth on it could not be distinguished, a model was developed describing the total gas exchange of mungbean and microbial growth, enabling the calculation of mungbean gas exchange in MA packages. A method was derived enabling the simultaneous measurement of metabolic gas exchange and the resistance to gas diffusion. Results of these measurements showed limitations to experimental setups using headspace techniques, and indicated that optimal O ₂ concentrations are very likely limited to a specific temperature. Measurements on acetaldehyde and ethanol confirm the prediction of the models describing fermentative CO ₂ production, and show that fermentation is not limited to low O ₂ concentrations.

Optimal gas conditions for storage of produce are not fixed values but change with temperature and, more important, also during the storage period. Using fixed gas conditions for long term storage, this could lead to problems and the loss of the stored produce. Interactive storage facilities, responding to physiology of the stored material, will help to reduce this risk. Processes that should be quantified are energy metabolism and fermentation rates. Parameters related to these processes are gas exchange rates and acetaldehyde and ethanol production. For the calculation of energy production the gas exchange models described within this thesis could be used. ATP fluxes, in combination with maintenance requirements, very likely help to understand the tolerance of plant tissues to low oxygen conditions.