27.1 Introduction Modified atmosphere (MA) techniques for horticultural products are based on the principle that manipulating or controlling the composition of the surrounding atmosphere affects the metabolism of the packaged product. By creating favourable conditions, quality decay of the product can be inhibited. The different MA techniques come with different levels of control to realise and/or maintain the composition of the atmosphere around the product. Passive MA packaging (MAP), as an extreme, relies solely on the metabolic activity of the packaged product to modify and subsequently maintain the gas composition surrounding the product. Temperature has a major effect on the rates of all processes involved in establishing the gas conditions in MAP (rates of gas exchange by the product and rates of diffusion through the packaging materials) and also on the rates of all metabolic processes that will inevitably lead to deterioration of the product and finally death. Ideally, steady state gas conditions should be obtained that, from the point of retaining quality, are optimal for the product packed. The time needed for a package to reach a steady state is extremely important as only from that moment in maximum benefit from MA being realised. Depending on conditions, the time to reach a steady state could theoretically outlast the shelf life of the packaged product. Given the ubiquitous role of temperature in MAP, success or failure of the ultimate MA package for a certain product largely depends on the level of integral temperature control from the moment of packing up to the moment of opening the package by the consumer. In logistic chains without integral temperature control, the application of MAP is often a waste of time, money and produce. In spite of the important role of temperature in MAP, most MAP research trials are performed at constant temperatures, at temperatures often close to what 27 MAP performance under dynamic temperature conditions M.L.A.T.M. Hertog, Katholieke Universiteit Leuven, Belgium is known as the optimum storage temperature for the product under study. No extensive literature data is available on monitoring MAP in terms of temperature, gas conditions and product quality throughout a logistic chain. Without such a complete set of data it is difficult, if not impossible, to know why a certain MAP design failed. This could, for instance, be due to a direct temperature effect on the product’s metabolism, or due to an indirect effect through a failure to establish the intended steady state gas conditions (too high or too low), or an unfortunate combination of other factors like leakage or issues related to product quality (maturity, microbial load, etc.). This chapter will focus on the effects of dynamic temperature conditions on the performance of MAP. First of all it will discuss how to define MAP performance; when MAP can be regarded as being successful and how this can be measured. Subsequently it will discuss what risks are involved in MAP and how these risks are affected by a lack of integral temperature control in a logistic chain. This chapter will conclude with a discussion of several simple strategies to maximise MAP performance, making the best of MAP given the limited resources available. The different aspects discussed in this chapter are illustrated using simulation results from a fully dynamic MA model 12 using realistic settings for both film and product characteristics. 27.2 MAP performance The first question to answer when discussing MAP performance is how MAP performance should be defined. The aim of MAP is to inhibit retardation of product quality, the means employed to reach this aim is the application of certain optimal MA temperature and gas conditions. To grade the performance of MAP one can test whether the aim was reached (in terms of product quality) or whether the means were employed correctly (in terms of temperature or gas conditions). If life were simple these two measures would be interchangeable, as they would be strongly correlated to each other. From a technical point of view, tracing and tracking gas conditions and temperature in the logistic chain is much easier than tracing and tracking those product properties responsible for the overall product quality. However, assessing the benefits and losses in terms of product quality gives much more insight than just the observation of MA conditions getting below or above their target levels. The question that should always be asked is how possible deviations in temperature or gas condition affect the quality and keeping quality. Product quality gives static information on the status of the product at a certain moment, for instance at the point of sale. Keeping quality provides dynamic information on how long a product can be stored, kept for sale, transported to distant markets or remain acceptable to the consumer. A wide range of equipment is available to monitor temperature throughout a logistic chain. Given that most MA packages are relatively small consumer packs and given the potentially large spatial and temporal variation in 564 Novel food packaging techniques temperature within cold stores and truckloads, there is a need to measure temperature at the level of the individual packs. Cheap versatile time temperature indicators (TTI) have been developed to give an indication of the temperature history to which individual packs have been exposed (See chapter 6). Even though these TTIs can give an indication of temperature abuse somewhere in the chain, they are not intended to reconstruct a complete temperature history and, therefore, cannot be expected perfectly to explain the resulting product quality. To give an example, a TTI will not discriminate between one week’s storage at 4oC disrupted by either 12 hours of continuous 12oC or six two-hour periods at 12oC. However, for the packed product this might make a difference, especially as the product needs time to heat up. With 12 hours of continuous 12oC the product will actually be at 12oC for part of that time. Exposed to the six two- hour periods of 12oC it depends on the time in between the warm periods how warm the product eventually will get. As a consequence, the two identical TTI readings from this example, can relate to two completely different qualities in the final product. Also the order of imposed temperatures will not make a difference to a TTI reading. However, for product quality, the order of the subsequent temperatures the product was exposed to might make a difference. For instance, pre-climacteric fruit generally responds less vigorously to temperature than the same fruit in its climacteric stage. With the effect of temperature on fruit physiology depending on the physiological stage of the fruit, two comparable temperature profiles (in terms of the total temperature sum) can have different effects in terms of product quality as this depends on the timing of the temperature relative to the physiological development of the fruit. The other important aspects of the established MA conditions are the gas conditions, which are inextricably related to temperature. As for temperature, several indicators have been developed to monitor oxygen (O 2 ) and carbon dioxide (CO 2 ) in individual packages. 16 As with TTIs, these gas indicators give only an indicative value. The potential strength of the different types of indicators arises from their combined application where information on temperature and gas conditions together can give a better indication of the realised MA conditions in individual MA packs. However, defining MAP performance by the realised MA conditions in terms of temperature and gas conditions is only an indirect measure. The ultimate unambiguous measure of the success of MAP is the final quality of the product. Some aspects of product quality can be related to volatiles produced by the product (ethylene as a measure of ripening stage, specific volatiles produced during spoilage or anaerobic conditions, etc.). This opens the door to adding product specific indicators to the range of indicators already available, resulting in the type of integrated freshness indicators as described in Chapter 7. Such freshness indicators might come close to giving a good evaluation of MAP performance incorporating several aspects of product quality into the equation. However, other aspects of product quality might never lend themselves to measurement in this way. MAP performance under dynamic temperature conditions 565 In spite of the importance of product quality as the ultimate determinant of MAP performance, this chapter will mainly focus on the effect of dynamic temperature conditions on the gas conditions developing inside MAP. Most of this is ruled by relatively simple physics. The link to product quality will be made when possible, but given the vast range of products and their different ways of responding to the applied MA conditions, 2, 11 no simple rules can be laid down on how dynamic temperature conditions will affect the quality of an MA packed product. For this, product specific knowledge is required on how product physiology responds to surrounding gas and temperature conditions in relation to the product at its own developmental stage. For now, one should be made aware that MAP performance is determined by more than just temperature and/ or the established gas conditions. 27.3 Temperature control and risks of MAP Like most techniques, MAP comes with a number of potential risks that largely depend on the level of integral temperature control in a logistic chain. 27.3.1 Low oxygen Generally, MAP is designed to create low levels of O 2 that give maximum benefit by suppressing the metabolism without getting into the range of O 2 levels that might induce fermentation. The critical O 2 level at which fermentation starts to occur is defined as the fermentation threshold. 18 The O 2 level in the package is the resultant of the influx through the package and consumption by the product. Both processes depend on temperature. O 2 consumption by the product generally increases much faster with increasing temperature (3- to 10-fold from 0–15oC 11 ) rather than the permeance of the packaging material (2- to 3-fold from 0–15oC 9 ). As a result, the steady state O 2 levels in the pack will decrease with increasing temperature. The O 2 level in a MA package designed to operate just above the fermentation threshold will, as a result of an increase in temperature, drop below this fermentation threshold; the product will start to ferment resulting in the development of off-odours and off- flavours. To make life more complicated, the fermentation threshold is not a constant but can vary with temperature. 1, 3, 18 When MA packed blueberries are exposed to a temperature increase, the drop in O 2 level is combined with an increase in fermentation threshold resulting in very little scope before anaerobic conditions are reached. Polymeric packaging materials that have the same responsiveness to temperature as the packed product can prevent induction of anaerobic conditions following increased temperature. In such cases an increase in O 2 consumption rate is counteracted by exactly the same increase in O 2 influx through the packaging material with the steady state gas conditions becoming independent of temperature. One such example was described for capsicums packed using 566 Novel food packaging techniques LDPE film. 7 One can argue whether a temperature-independent atmosphere inside the package is important in its own right. The aim of MAP is to retain quality. With constant gas conditions at increasing temperatures, respiration rate and the rate of quality decay will still increase due to the increased temperature. The O 2 levels in MA packages that make use of perforated films are even more sensitive to changes in temperature, as diffusion through the holes (i.e. diffusion through a barrier of standing air) is almost independent of temperature. An increase in temperature will induce increased O 2 consumption by the product without inducing a substantial increased influx through the packaging material, resulting in a fast drop of the steady state O 2 levels. 27.3.2 High carbon dioxide Besides reducing O 2 levels in MAP, CO 2 levels are increased to further inhibit the product’s metabolism. 2 High CO 2 levels also inhibit decay by suppressing the growth of microbes, although sometimes the CO 2 levels needed to suppress microbial growth exceed the tolerance levels of the vegetable produce packaged. 4, 6 This identifies another dilemma in controlling the gas conditions in MAP. For most polymeric packaging films the permeance for CO 2 is 2- to 10-fold higher than for O 2 , 9 under aerobic conditions O 2 depletes much faster than CO 2 will accumulate. Assuming a respiratory quotient of 1 and a steady state O 2 level of 2kPa, the maximum achievable steady state CO 2 level varies between 2 and 9kPa depending on the film material. To achieve higher steady state CO 2 levels without inducing fermentation, microperforated films should be used that have comparable permeances for O 2 and CO 2 . When using microperforated films, O 2 will deplete about as fast as CO 2 accumulates, such that the sum of O 2 and CO 2 partial pressure remains around 20kPa. A microperforated MA package designed for 2kPa O 2 can therefore generate CO 2 levels of around 18kPa. For soft fruit like strawberries, these high CO 2 levels are needed to prolong shelf- life. 8, 13 However, after prolonged storage at high CO 2 (>15 kPa) CO 2 injury becomes visible from tissue defects and fermentation off-flavours. 10, 14 When exposing MA packages to dynamic temperature conditions there is a direct risk of inducing fermentation and an added secondary risk of inducing CO 2 damage due to the accumulating fermentative CO 2 . Especially for microperforated packs where the permeance does not increase with temperature, the risk of inducing fermentation and consequently the accumulation of high CO 2 levels is much larger. Scavengers to constrain the accumulation of CO 2 (Chapter 3) might limit the secondary risk of CO 2 damage but cannot prevent the direct risk of inducing fermentation. 27.3.3 High humidity With horticultural products generally consisting of up to 90% water and with their economic value often determined by the saleable weight of the crop, moisture loss needs to be limited under all conditions. Depending on how and MAP performance under dynamic temperature conditions 567 for how long horticultural products are stored, they can easily lose up to 5% or more of their harvested weight before they reach the consumer. Generally, MAP films, either perforated or not, are relatively impermeable to water vapour and therefore quickly generate high humidity levels in the package atmosphere close to saturation. With dropping temperatures the saturating vapour pressure drops as well and the colder air cannot continue to hold as much water vapour. Due to the extremely low water vapour permeance of most films, water vapour cannot leave the package fast enough, resulting in condensation in the package. This will happen with temperature fluctuations as small as 0.5oC. In the heat of harvest activities, there is often not enough capacity properly to cool the product before packing. Packing warm product in plastic film, either for MA purposes or as liners in carton boxes, followed by cool storage, also results in extreme condensation inside the package thus wetting the product. The high humidity levels generated inside MAP prevent excessive water loss from the product retaining product quality, but the presence of free water following temperature fluctuations creates favourable conditions for microbes to flourish and break down this same product quality. 27.4 The impact of dynamic temperature conditions on MAP performance As outlined in Chapter 16, different sources of variation interact with the performance of MA packages. In this chapter we discuss the effect of temperature variation over time, and how that can affect MA conditions and final product quality. To allow for some temperature flexibility, MAP should be designed to prevent those risks outlined in the previous sections (too low O 2 , too high CO 2 , too high humidity). The closer package atmospheres are targeted to what is feasible, the more likely temperature variation can induce these risks. How closely the theoretical ideal gas conditions can be approximated depends not only on the amount of temperature variation one wants to allow for but also on the amount of variation in other relevant aspects and on how temperature interacts with these. For instance, when aimed for O 2 levels are close to the fermentation threshold, depending on the variation in gas exchange rate, there is a risk that some of the packages result in O 2 levels dropping below the fermentation threshold. 5 Depending on the variation in the fermentation threshold itself and the variation in film permeability, tightness of seal, number of layers wrapped around the product, etc., the targeted safe gas conditions might need to be far removed from the theoretical ideal gas conditions. With the number of variables encountered in MA packaging it is difficult to give full coverage of all aspects of the impact of dynamic temperature profiles on MAP, as this strongly depends on the specifications of the package of interest. Some of the important aspects are now discussed using simulations of MA packaging of shredded lettuce. 568 Novel food packaging techniques The quality of shredded lettuce is often limited by browning of the cut edges. This can be controlled by packaging in <1% O 2 and 10% CO 2 atmos- pheres. 15, 17 Shredded lettuce is a product with a relative high respiration rate and a high responsiveness to temperature as expressed by the energy of activation of respiration (see Chapter 16). As a reference we simulated MAP of pre-cooled lettuce stored at a constant 4oC and packed in a polymeric bag with an energy of activation of about one-third of the lettuce itself (Fig. 27.1b and c). Steady state gas conditions (10kPa CO 2 and 1kPa O 2 ) are reached after about two and a half days of storage with O 2 levels reaching 2 kPa after one-day storage. The realised steady state gas conditions correspond to the targeted optimum values for shredded lettuce. When one realises that minimally processed products generally have a limited shelf-life, the two and a half days needed to establish steady state conditions is relatively long. For subsequent simulations an artificial dynamic temperature profile was created (Fig. 27.1a) consisting of one day at a constant 4oC followed by a two- day period of slow fluctuating temperature around 4oC and a subsequent one-day period of fast fluctuating temperature. After this, temperature was rapidly increased to a constant 12oC. Instead of assuming the lettuce to be pre-cooled, lettuce was assumed to be at room temperature when packed. As a result of packing warm lettuce, depletion of O 2 and accumulation of CO 2 was accelerated in comparison to the reference situation (Fig. 27.1b and c), the O 2 level of 2 kPa was reached only half a day after packing. Both O 2 and CO 2 show fluctuating levels in response to the fluctuating temperature of the environment. The fluctuations in O 2 and CO 2 follow the fluctuations in temperature after a short delay, as the product needs time to warm up and cool down. The larger the thermal mass and heat capacity of the product, the slower the product will respond to fluctuations in temperature. This explains why gas levels follow slow temperature fluctuations more clearly than they follow the fast temperature changes. Another reason why gas levels do not follow fast temperature changes is because of the void volume in the package, which buffers the change in gas conditions. The direction of the fluctuation in CO 2 level is the same as for temperature while the direction of the fluctuation in O 2 level is the opposite. As temperature increases, film permeance increases. However, the rate of O 2 consumption increases faster than the increase in film permeance resulting in dropping O 2 levels. With dropping O 2 levels fermentative CO 2 production increases resulting in increasing levels of CO 2 . During the period of fluctuating temperature the same average gas levels are reached as seen before. When temperature is increased to 12oC, the O 2 level drops to 0.5kPa while CO 2 accumulates up to 18kPa due to the fermentation induced. It will be clear that such an increase in temperature to 12oC when a package is designed to operate around 4oC is fatal for the packed product. Depending on the product such temperature increase might irreversibly affect product quality. Packing warm product has the advantage of rapidly establishing the targeted gas conditions. The downside is the induction of condensation as the warm MAP performance under dynamic temperature conditions 569 Fig. 27.1 Simulation results of MA packed shredded lettuce stored at a constant 4oC or at dynamic temperature conditions. (a) Temperature profile used for the dynamic temperature conditions; air temperature (——) and product temperature ( ). (b) O 2 levels observed in the package during different simulation runs. (c) CO 2 levels observed in the package during different simulation runs. (d) Condensation formed during dynamic temperature conditions. The following simulations are depicted in (b) and (c): reference simulation of pre-cooled lettuce packed in polymeric film and stored at a constant 4oC (——), lettuce packed warm using polymeric film and stored at dynamic temperature conditions ( ), lettuce packed warm and stored at dynamic temperature conditions but with a reduced void volume ( . . . . . ), lettuce packed warm using microperforated polymeric film and stored at dynamic temperature conditions ( . – . – . – ). The boxes in (b) and (c) contain an enlargement of what is happening during the period with fluctuating temperatures. 570 Novel food packaging techniques product evaporates more water than the cold air can contain, quickly oversaturating the air with an excess water condensating on the inside of the cold packaging material (Fig. 27.1d). During the subsequent period, con- densation slowly disappears again by evaporation and diffusion through the film. With fluctuating temperatures the amount of condensate fluctuates as well. Once temperature is increased to 12oC there is a fast drop in the amount of condensate. These relative fast changes are due to changes in the air saturation levels for water vapour as a function of temperature. This example shows that condensation can be rapidly induced but once present is hard to remove without increasing temperature again. When the void volume in the package is eliminated (Fig. 27.1b and c) steady state gas conditions are rapidly realised within half a day. Because of the warm lettuce, the CO 2 level peaks to initially extremely high levels, rapidly disappearing when the product cools down. By reducing the void volume we have removed the buffering capacity of the system as a consequence of which the gas levels respond much more vigorously to the fluctuating temperature and also become more sensitive to fast fluctuations. When temperature is increased to 12oC, the increase in CO 2 is much faster than before. When the film is replaced by a microperforated material, permeance of the packaging film has become almost independent of temperature. The resulting gas conditions are now different (Fig. 27.1b and c) with O 2 going towards 3kPa and CO 2 continuing to increase with time. The reason for not reaching steady state conditions is the relatively much lower permeance for CO 2 as compared to the permeance for O 2 . Therefore the steady state conditions for CO 2 are at much higher CO 2 levels than before, which takes more time and the MA package never reaches this situation. Because of the temperature independency of film permeance the fluctuations in O 2 levels respond vigorously to changes in temperature. The final temperature increase to 12oC results in a drop of O 2 to 1kPa and an increase of CO 2 towards 40–50kPa. This increase is clearly the result of fermentative CO 2 production that, due to the low permeance for CO 2 is trapped inside the package. As the accumulating CO 2 has an inhibitive effect on the respiration of lettuce, O 2 consumption is inhibited, resulting in a subsequent slight increase of the O 2 level. The outlined simulations were focused on a single average MA pack. When the dynamic temperature condition is applied to a batch of MA packages, each prepared package will differ slightly from another. Given that biological variance is the most variable parameter, we simulated a batch of 500 packages assuming 25% variation on product respiration rates, and 10% variation on packed product weight and film thickness (Fig. 27.2). The simulation result clearly shows the effect of variation in MA design parameters on the resulting MA gas conditions. At the same time it shows that variation in MA gas conditions depends on time and temperature. As, depending on the respiration rates, some packages establish MA conditions faster than others, initially a large variation in MA gas conditions is observed. Some packages reached a level of 2kPa O 2 within three hours after packing while others took two days to reach MAP performance under dynamic temperature conditions 571 this stage. By reducing the void volume, packing warm product, or flushing the package with nitrogen, the process of establishing MA conditions can be facilitated reducing the initial large variation in MA gas conditions. The variation in O 2 levels is generally much smaller than the variation in CO 2 levels, especially when the temperature increase to 12oC induces fermentation. Under these conditions the high CO 2 levels in some of the packs will induce CO 2 injury. Controlling temperature in such a way that none of the packs develop fermentation would keep the variation in CO 2 levels within limits. 27.5 Maximising MAP performance From the simulations in the previous section it became clear that it is of the utmost importance to prevent all sources of variation, whether that is temperature variation (time but also spatial variation), variation in the product (maturity differences causing variation in respiration rate or variation in the amount of product packed), or variation in the homogeneity of the package itself (variation in thickness, perforations, layers of wrapping, tightness of seal, etc.). Biological variation tends to average itself out when large enough batches of product are packed. The variation between consumer MA packs containing a limited amount of product will be much larger than variation between MA Fig. 27.2 Simulation results of 500 MA packages of shredded lettuce packed using polymeric film stored at dynamic temperature conditions (Fig. 27.1a). The average CO 2 (——) and O 2 levels ( ) are plotted together with their 95% confidence intervals ( . . . . . ). 572 Novel food packaging techniques packed pallets containing a large amount of product per pallet. So increasing the size of MA packages can cope with within-batch variation. Potentially, there is also a large variation related to the maturity of the packed product during the course of the season. As a consequence, early harvested product might have different packaging needs from product harvested later in the season. Ideally, the design of a MA pack is adapted during the season to cope with these changes in maturity. Fine-tuning the design of MA packages to these changing needs during the season can theoretically be done by relatively simple measures as long as one knows what the changing needs of the product are. Close co-operation between product and packaging experts is needed to develop guidelines for the horticultural packaging companies. Variation in the homogeneity of the physical package (variation in thickness, perforations, layers of wrapping, tightness of seal, etc.) is a technical issue that is relatively easy to control during the production process by appropriate quality control. To enable rapid establishment of the intended MA conditions several simple techniques can be applied such as gas flushing the package before sealing. Although this is the most expensive technique, it can establish steady state gas reliably and instantaneously. Packing of warm fruit is the simplest way but comes with the risk of inducing lots of free water in the pack. Depending on how vulnerable the product is to microbial breakdown this might not be an option. Reducing the void volume is the third way of speeding up the process of establishing steady state gas conditions. However, this is not only speeding up the initial process of establishing steady state gas conditions but is increasing the overall responsiveness of the package allowing it rapidly to follow any temperature fluctuations in the logistic chain. Temperature variation can be minimised only by an integral temperature control throughout the whole logistic chain from field to table. It is of the utmost importance to involve all partners in the chain in this integral temperature control as any temperature abuse might nullify the efforts of all other partners. In the end, the success of a chain is determined by the weakest link in the chain. When designing MAP for a certain product one should consider whether the potential benefits are worth the possible risks of a lack of temperature control. If this is questionable, one might consider designing a safe MA system by designing it for the highest temperature likely to be encountered. Although this approach does not utilise the maximum benefits it rules out all associated risks. In the end, MAP can only be successful when good temperature control can be guaranteed. 27.6 Future trends The eventual success of MA depends on temperature control between the moment of packing and the moment of opening of the package by the consumer. Instead of relying solely on one’s gut feeling when optimising MAP, a MAP model to simulate a package going through a logistic chain will give insight into MAP performance under dynamic temperature conditions 573 the strong and weak parts of that chain in terms of temperature control. 12 It will make clear which parts of the chain are responsible for the largest quality losses of the packaged product and need improvement. It enables the optimisation of a whole chain considering the related costs and benefits. To operate such a model, information is needed on temperature, O 2 and CO 2 dependencies of gas exchange and on temperature dependency of film permeance. With regard to the temperature effect on the oxidative respiration of different fruits and vegetables there is some data available. 11 Information on fermentation and on the effects of O 2 and CO 2 on gas exchange is much more fragmentary. This makes it almost impossible to identify at what temperature anaerobic conditions are going to be induced. Also a good database on permeance of packaging films that includes their temperature dependency is lacking. Before a new film can be used for MAP its temperature characteristics need to be identified at temperatures relevant to MAP (0 25oC). To be able to bring MAP to the next level and to predict what the effect of certain dynamic temperature conditions is on a particular MAP design it is vital to establish such databases on product and film characteristics. Without this elementary knowledge, MAP will remain at the level of trial and error. Ultimately, any temperature variation in the logistic chain should be ruled out. Meanwhile, technical solutions like temperature sensitive films are emerging to cope with some of the existing dynamic temperature conditions. 27.7 References 1. BEAUDRY R M, CAMERON A C, SHIRAZI A and DOSTAL-LANGE D L, ‘Modified- atmosphere packaging of blueberry fruit: effect of temperature on package O 2 and CO 2 ’, J. Amer. Soc. Hort. Sci., 1992 117 436–41. 2. BEAUDRY R M, ‘Effect of O 2 and CO 2 partial pressure on selected phenomena affecting fruit and vegetable quality’. Postharvest Biology & Technology, 1999 15 293–303. 3. BEAUDRY R M and GRAN C D, ‘Using a modified-atmosphere packaging approach to answer some postharvest questions: Factors affecting the lower oxygen limit.’ Acta Hort. 1993 362 203–12. 4. BENNIK M H J, Biopreservation in modified atmosphere packaged vegetables, Thesis Wageningen Agricultural University. ISBN 90-5485- 808-7, 1997. 5. CAMERON A C, PATTERSON B D, TALASILA P C and JOLES D W, ‘Modeling the risk in modified-atmosphere packaging: a case for sense-and-respond packaging’, in: Blanpied. G D, Bartsch, J A and Hicks, J R (eds), Proc. 6th Intl. Controlled Atmosphere Research Conference, Ithaca NY, 1993. 6. CHAMBROY Y, GUINEBRETIERE M H, JACQUEMIN G, REICH M, BREUILS L and SOUTY M, ‘Effects of carbon dioxide on shelf-life and post harvest decay of strawberries fruit’, Sciences des Aliments, 1993 13 409–23. 7. CHEN X Y, HERTOG M L A T M and BANKS N H, ‘The effect of temperature on 574 Novel food packaging techniques gas relations in MA packages for capsicums (Capsicum annuum L., cv. Tasty): an integrated approach’. Postharvest Biology & Technology, 2000 20 71–80. 8. COLELLI G and MARTELLI S, ‘Beneficial effects on the application of CO 2 - enriched atmospheres on fresh strawberries (Fragaria ananassa Duch.).’ Adv. Hortic. Sci., 1995 9 55–60. 9. EXAMA A, ARUL J, LENCKI RW, LEE L Z and TOUPIN C, ‘Suitability of plastic films for modified atmosphere packaging of fruit and vegetables.’ J. Food Sci., 1993 58 1365–70. 10. GIL M I, HOLCROFT D M and KADER A A, ‘Changes in strawberry anthocyanins and other polyphenols in response to carbon dioxide treatments.’ J. Agric. Food Chem. 1997 45 1662–7. 11. GROSS, K C, WANG C Y and SALTVEIT M E, ‘The Commercial Storage of Fruits, Vegetables, and Florist and Nursery Crops.’ An Adobe Acrobat pdf of a draft version of the forthcoming revision to US Department of Agriculture, Agriculture Handbook 66 on the website of the USDA, Agricultural Research Service, Beltsville Area <http://www.ba.ars. usda.gov/hb66.html> (November 8, 2002). 12. HERTOG M L A T M, PEPPELENBOS H W, TIJSKENS L M M and EVELO R G, ‘Modified atmosphere packaging: optimisation through simulation’, in: Gorny, J R (ed.), Proc. 7th Intl. Controlled Atmosphere Research Conference, Davis CA, 1997. 13. HERTOG M L A T M, BOERRIGTER H A M, VAN DEN BOOGAARD G J P M, TIJSKENS L M M and VAN SCHAIK A C R, ‘Predicting keeping quality of strawberries (cv. ‘Elsanta’) packed under modified atmospheres: an integrated model approach.’ Postharvest Biology & Technology, 1999 15 1–12. 14. KE D, GOLDSTEIN L, O’MAHONY M and KADER A A, ‘Effects of short-term exposure to low O 2 and high CO 2 atmospheres on quality attributes of strawberries.’ J. Food Sci. 1991 56 50–4. 15. LOPEZ-GALVEZ G, SALTVEIT M E and CANTWELL M, ‘The visual quality of minimally processed lettuce stored in air or controlled atmospheres with emphasis on romaine and iceberg types.’ Postharvest Biology & Technology, 1996 8 179–90. 16. SMOLANDER M, HURME E and AHVENAINEN R, ‘Leak indicators for modified-atmosphere packages.’ Trends in Food Science & Technology, 1997 8 101–6. 17. SMYTH A B, SONG J and CAMERON A C, ‘Modified atmosphere packaged cut iceberg lettuce: effect of temperature and O 2 partial pressure on respiration and quality.’ J. Agric. Food Chem. 1998 46 4556–62. 18. YEARSLEY C W, BANKS N H and GANESH S, ‘Temperature effects on the internal lower oxygen limits of apple fruit.’ Postharvest Biology & Technology, 1997 11 73–83. MAP performance under dynamic temperature conditions 575