Effects of reduction in beef surface water activity on the survival of Salmonella Typhimurium DT104 during heating

Aim: To investigate the influence of reducing beef surface water activity (a_w) on the survival of Salmonella Typhimurium DT104 during heating. Methods and Results: Beef discs were surface inoculated with S. Typhimurium DT104 and either untreated or dried to achieve surface a_w values of 0·95, 0·85 and 0·70. The samples were vacuum packed, heat‐treated at 60°C and removed at predetermined times. The inactivation curves were influenced by a_w and treatment time. Biphasic inactivation curves were observed for S. Typhimurium DT104 heat‐treated on beef samples with altered a_w values, which were characterized by an initial decline in cell numbers at commencement of heating followed by a much slower rate of inactivation during the remaining treatment period. Point estimates of the heating time required to achieve a 1 log reduction on beef surfaces with a_w of 0·99, 0·95, 0·85 and 0·70 were 0·5, 1·55, 11·25 and 17·79 min, respectively. Conclusions: A decrease in beef surface a_w can substantially enhance the survival of S. Typhimurium DT104 after heating. Significance and Impact of the Study: Caution needs to be taken using dry air as a decontamination method as this may rapidly decrease product surface and pathogen a_w values resulting in enhanced survival.     

The water content/water activity relationship of cured tobacco and water relations of associated spoilage fungi

The relationship between moisture content and water activity (a_w) in cured tobacco was significantly influenced by sugar content. Overall, high sugar tobaccos such as Oriental and Virginia had a higher moisture content at any given water activity compared to low sugar tobaccos such as Burley. Virginia and Burley were both predominantly colonised by Aspergillus and Penicillium spp. Of these, about 80% of isolates could germinate at between 0·75 and 0·85 a_w, equivalent to moisture contents of between 18% and 24% in Burley and between 22% and 31% in Virginia. Growth of the dominant Aspergillus and Penicillium spp. was much slower on Virginia and Burley tobacco extract than on malt extract agars over the range 0.85 to 0.98 a_w. For some species the optimum a_w for growth on tobacco extract medium was altered from that on the richer malt extract agar and for some there was also a significant difference in growth between Virginia and Burley extract agars. The mould-free storage periods for five different tobacco types was influenced by a_w. Visible moulding occurred within 7–14 days at 0.85–0.90 a_w but only after about six months at 0.70–0.75 a_w. There were some differences in rate of moulding between tobacco types as well as in the range of fungi isolated at different a_w storage levels.     

Effect of temperature and water activity on growth and ochratoxin A production boundaries of two Aspergillus carbonarius isolates on a simulated grape juice medium

Aims: To develop and validate a logistic regression model to predict the growth and ochratoxin A (OTA) production boundaries of two Aspergillus carbonarius isolates on a synthetic grape juice medium as a function of temperature and water activity (a_w). Methods and Results: A full factorial design was followed between the factors considered. The a_w levels assayed were 0·850, 0·880, 0·900, 0·920, 0·940, 0·960, 0·980 and the incubation temperatures were 10, 15, 20, 25, 30, 35 and 40°C. Growth and OTA production responses were evaluated for a period of 25 days. Regarding growth boundaries, the degree of agreement between predictions and observations was >99% concordant for both isolates. The erroneously predicted growth cases were 3·4–4·1% false‐positives and 0·7–1·4% false‐negatives. No growth was observed at 10°C and 40°C for all a_w levels assayed, with the exception of 0·980 a_w/40°C, where weak growth was observed. Similarly, OTA production was correctly predicted with a concordance rate >98% for the two isolates with 0·7–1·4% accounting for false‐positives and 2·0–2·7% false‐negatives. No OTA production was detected at 10°C or 40°C regardless of a_w, and at 0·850 a_w at all incubation temperatures. With respect to time, the OTA production boundary shifted to lower temperatures (15–20°C) as opposed to the growth boundary that shifted to higher temperature levels (25–30°C). Using two literature datasets for growth and OTA production of A. carbonarius on the same growth medium, the logistic model gave one false‐positive and three false‐negative predictions out of 68 growth cases and 13 false‐positive predictions out of 45 OTA production cases. Conclusions: The results of this study suggest that the logistic regression model can be successfully used to predict growth and OTA production interfaces for A. carbonarius in relation to temperature and a_w. Significance and Impact of the Study: The proposed modelling approach helps the understanding of fungal‐food ecosystem relations and it could be employed in risk analysis implementation plans to predict the risk of contamination of grapes and grape products by A. carbonarius.     

Effect of temperature and water activity on spore germination and mycelial growth of three fungal biocontrol agents against water hyacinth (Eichhornia crassipes)

Aims: To determine the effect of water activity (a_w = 0·880–0·960) and temperature (15–35°C) on the percentage of viable conidia and mycelial growth of three biocontrol agents effective against water hyacinth in Mali: Alternaria sp. isolate Mlb684, Fusarium sacchari isolate Mln799 and Cadophora malorum isolate Mln715. Methods and Results: The fungi were grown in vitro on plates containing potato dextrose agar medium at different a_w values (glycerol being added to adjust the a_w). The percentage of viable conidia and radial growth rate decreased with decreasing water activity. Statistical analysis showed a significant effect of a_w, temperature and the a_w × temperature interaction on mycelial growth (P < 0·0001). Water activity emerged as the factor exerting the greatest influence. Differences were observed between the fungi tested, the C. malorum appearing more tolerant to low a_w and the F. sacchari more tolerant to high temperature (35°C). Growth models predicting the combined effect of a_w and temperature were developed and response surfaces generated, showing fairly good agreement with the experimental values. Conclusions: Our results confirm the previous finding that a_w has a greater influence than temperature on fungal growth. Under most conditions, variation of environmental factors has a detrimental influence on the percentage of viable conidia and mycelial growth rate of fungal isolates. Significance and Impact of the Study: The developed models may contribute to predicting the best environmental conditions for use of these fungi as effective biocontrol agents against water hyacinth.     

Effect of environmental factors on tenuazonic acid production by Alternaria alternata on soybean-based media

Aims: To determine the effects of water activity (a_W; 0·995–0·90), temperature (5, 18, 25 and 30°C), time of incubation (7–35 days) and their interactions on tenuazonic acid (TA) production on 2% soybean‐based agar by two Alternaria alternata strains isolated from soybean in Argentina. Methods and Results: TA production by two isolates of A. alternata was examined under interacting conditions of a_W, temperature and time of incubation on 2% soybean‐based agar. Maximum TA production was obtained for both strains at 0·98 a_W, but at 30 and 25°C for the strains for RC 21and RC 39, respectively. The toxin concentration varied considerably depending on a_W, temperature, incubation time and strain interactions. TA was produced over the temperature range from 5 to 30°C and a_W range from 0·92 to 0·995, however at 5 and 18°C little TA was produced at a_W below 0·94. Contour maps were developed from these data to identify areas where conditions indicate a significant risk for TA accumulation. Conclusions: The optimum and marginal conditions for TA production by A. alternata on soybean‐based agar were identified. The results indicated that TA production by A. alternata is favoured by different temperatures in different strains. Significance and Impact of the Study: Data obtained provide very useful information for predicting the possible risk factors for TA contamination of soybean as the a_W and temperature range used in this study simulate those occurring during grain ripening. The knowledge of TA production under marginal or sub‐optimal temperature and a_W conditions for growth are relevant as improper storage conditions accompanied by elevated temperature and moisture content in the grain can favour further mycotoxin production and lead to reduction in grain quality.     

Survival of alginate-entrapped cells of Azospirillum lipoferum during dehydration and storage in relation to water properties

Survival of alginate-entrapped cells of Azospirillum lipoferum was studied during dehydration using a dry air stream and during prolonged storage at various constant water activity values (a_w). During the drying operation, the viability loss remained almost constant from the initial water content to 0.35 g water/g dry weight (DW) corresponding to a 98.5% water removal, strongly increased until a water content of 0.25 g/g DW and then stopped until the end of the drying operational (final a_w 0.18). A water content of 0.25 g/g DW (a_w=0.55) corresponded to the critical point of the moisture sorption isotherm curve from which water became restricted to the dry material. A high drying rate (5 g/g DW per hour) was shown to be more detrimental for cell viability than a low drying rate (1.18 g/g DW per hour). When the product was stored in a closed chamber with a regulated a_w (0.23), the number of living cells decreased during a short period (less than 15 days) corresponding to the product a_w stabilization, and then remained constant for more than 150 days. In addition, cell survival during storage was not affected by a_w values in the range 0–0.55. Above a_w=0.55, the higher the a_w and the storage duration, the lower the residual survival percentage. The influence of the drying and storage conditions on the cell death rate are discussed with regard to both the mechanisms generally involved in viability loss and the hydration properties of water. Correspondence to: A. Pareilleux     

Modelling the effect of water activity and temperature on growth rate and aflatoxin production by two isolates of Aspergillus flavus on paddy

Aims: This study was conducted to characterize the growth of and aflatoxin production by Aspergillus flavus on paddy and to develop kinetic models describing the growth rate as a function of water activity (a_w) and temperature. Methods and Results: The growth of A. flavus on paddy and aflatoxin production were studied following a full factorial design with seven a_w levels within the range of 0·82–0·99 and seven temperatures between 10 and 43°C. The growth of the fungi, expressed as colony diameter (mm), was measured daily, and the aflatoxins were analysed using HPLC with a fluorescence detector. The maximum colony growth rates of both isolates were estimated by fitting the primary model of Baranyi to growth data. Three potentially suitable secondary models, Rosso, polynomial and Davey, were assessed for their ability to describe the radial growth rate as a function of temperature and a_w. Both strains failed to grow at the marginal temperatures (10 and 43°C), regardless of the a_w studied, and at the a_w level of 0·82, regardless of temperature. Despite that the predictions of all studied models showed good agreement with the observed growth rates, Davey model proved to be the best predictor of the experimental data. The cardinal parameters as estimated by Rosso model were comparable to those reported in previous studies. Toxins were detected in the range of 0·86–0·99 a_w with optimal a_w of 0·98 and optimal temperature in the range of 25–30°C. Conclusions: The influences of a_w and temperature on the growth of A. flavus and aflatoxin production were successfully characterized, and the models developed were found to be capable of providing good, related estimates of the growth rates. Significance and Impact of the Study: The results of this study could be effectively implemented in minimizing the risk of aflatoxin contamination of the paddy at postharvest.     

Viability of dry Trichoderma harzianum spores under storage

Spores of the potential biocontrol agent Trichoderma harzianum P1 were prepared without (M1) and with heat shock (40?°C for 90?min) after fermentation (M2), filtered into a paste and dried over silica gel. M1 and M2 exhibited high viability (55%) and similar initial trehalose contents (4.0 and 5.4%, respectively) after slow drying. No significant differences in viability were found between treatments during storage for 110 days under different temperatures, T (8, 33 and 42?°C) and water activities, a _w (0.03, 0.33 and 0.75). Viability of spores, after storage at a _w =0.03 were 100 and 70% for 8 and 33?°C, respectively. During storage, decrease in trehalose content and viability was faster at a _w =0.75 and 42?°C. Loss of viability was modeled by a first order kinetic model depending on 1/T and a _w . M2 (with heat shock) showed slightly higher trehalose contents than M1 which resulted in 100% viability after 52 days at 8?°C.     

Modelling time to growth of Escherichia coli as a function of water activity and undissociated lactic acid

Aims: To model the effect of water activity (a_w) and concentration of undissociated lactic acid (HLac) on the time to growth (TTG) and the growth/no growth boundary of acid‐adapted generic Escherichia coli, used as model organisms for Shiga toxin‐producing E. coli (STEC). Methods and Results: For each of two E. coli strains, the TTG in brain heart infusion broth at 27°C was estimated at 30 combinations of a_w (range 0·945–0·995) and concentration of HLac (range 0–6·9 mol m^?3) by using an automated turbidity reader. Survival analysis was used to develop a model predicting the TTG and the growth/no growth boundary. Conclusions: The present model can be used to predict the TTG and to indicate the growth/no growth boundary of acid‐adapted E. coli strains as a function of a_w and concentration of HLac. Significance and Impact of the Study: Fermented food products have been implicated as sources of STEC in several outbreaks. The study results are relevant for modelling of growth of STEC in fermented food and can be used in microbiological risk assessments or in the design and validation of food‐production processes.     

Production of the fungal biocontrol agent Epicoccum nigrum by solid substrate fermentation: effect of water activity on accumulation of compatible solutes

Epicoccum nigrum conidia were produced by solid fermentation on wheat grains (cv. Rendeveaux and Brigadier) at different water activities (a_w). Conidial production was highest at high a_w(0.996) than at reduced a_w (0.98). However, conidial production at reduced a_w was improved when the a_w of the substrate was adjusted with a mixture of glycerol and water. Maximum levels ofconidiation were 7–11 × 10⁶ conidia g⁻¹ grain. The a_w of the solid substrate affected the pattern of accumulation of compatible solutes in the conidia. Mannitol was the main polyol in all conidialtypes. However, the amounts of mannitol were higher in conidia produced at high a_w. At reduced a_w the conidia of E. nigrum accumulated moreglycerol, which is more efficient in the osmorregulation proccess than mannitol. Arabitol accumulated in low amounts, specifically in conidia produced at the lower a_w, on cv. Rendeveaux but not on cv. Brigadier. Trehalose was detected in higher amounts in cv. Rendeveaux than in cv. Brigadier, andthe amounts were higher in conidia produced at high a_w. A significant amount of endogenous solutes was detected in the washing liquid used for the separation of the conidia. This revised version was published online in June 2006 with corrections to the Cover Date.     

Impact of a Streptomyces (AS1) strain and its metabolites on control of Aspergillus flavus and aflatoxin B1 contamination in vitro and in stored peanuts

Six actinomycetes were isolated from peanuts in Egypt. Of these, a Streptomyces strain (AS1) was found in in vitro assays to inhibit directly or via secondary metabolites both germination and growth of Aspergillus flavus. Tests of the AS1 cells for direct control of A. flavus populations or aflatoxin B₁ (AFB₁) production on stored peanuts was unsuccessful over 14-day storage periods. However, crude extracts of AS1 metabolites at 50 and 100 ppm completely inhibited spore germination of conidia of A. flavus in vitro over 48 h. Comparison of solvents for extracting the metabolites showed that the ethyl acetate extract was most effective. This gave greater than 85% inhibition of mycelial growth at these concentrations at different water availabilities (water activity; a _w; 0.95, 0.92, and 0.89) and 25°C. Doses of 50, 200, and 500 ppm of AS1 metabolites significantly inhibited populations of A. flavus on stored peanuts at two water stress levels (0.90, 0.93 a _w) at 25°C over 14-day storage periods. The amounts of AFB₁ produced by A. flavus on peanuts stored at 0.90 a _w were significantly decreased by AS1 metabolites for only 7 days. However, at 0.93 a _w doses of 200 and 500 ppm significantly controlled AFB₁ accumulation in peanuts for 14 days.     

Effect of water activity adjusted with different solutes on growth and lactic acid production byLactobacillus helveticus

Effect of water activity (a _w) adjusted with NaCl or glycerol on the growth and metabolism ofLactobacillus helveticus var.pragensis at 40°C was demonstrated by growth curves (generation time) and changes in the degree of acidity of a milk medium expressed as lactic acid content. NaCl-regulateda _w of 0.970 inhibited growth and acid production completely. The same, glycerol-regulateda _w of 0.970 increased the generation time only from 33 min (a _w=0.994) to 67 min which was less than the generation time at the highera _w of 0.982 adjusted with NaCl (103 min). Glycerol-regulateda _w of 0.970 did not decrease the acid production (2.6% lactic acid). Ata _w of 0.951, the acid production was decreased by 39% compared with the values found in milk media with the original, unadjusteda _w of 0.994, after the same time of incubation. Media witha _w adjusted to the same values with NaCl or glycerol do not influence the growth and acid production ofL. helveticus in the same way. In contrast to glycerol, NaCl has a strong effect.     

Modelling the effect of temperature, water activity and pH on the growth of Serpula lacrymans

Aims: To predict the risk factors for building infestation by Serpula lacrymans, which is one of the most destructive fungi causing timber decay in buildings. Methods and Results: The growth rate was assessed on malt extract agar media at temperatures between 1·5 and 45°C, at water activity (a_w) over the range of 0·800–0·993 and at pH ranges from 1·5 to 11·0. The radial growth rate (μ) and the lag phase (λ) were estimated from the radial growth kinetics via the plots radius vs time. These parameters were then modelled as a function of the environmental factors tested. Models derived from the cardinal model (CM) were used to fit the experimental data and allowed an estimation of the optimal and limit values for fungal growth. Optimal growth rate occurred at 20°C, at high a_w level (0·993) and at a pH range between 4·0 and 6·0. The strain effect on the temperature parameters was further evaluated using 14 strains of S. lacrymans. The robustness of the temperature model was validated on data sets measured in two different wood‐based media (Quercus robur L. and Picea abies). Conclusions: The two‐step procedure of exponential model with latency followed by the CM with inflection gives reliable predictions for the growth conditions of a filamentous fungus in our study. The procedure was validated for the study of abiotic factors on the growth rate of S. lacrymans. Significance and Impact of the Study: This work describes the usefulness of evaluating the effect of physico‐chemical factors on fungal growth in predictive building mycology. Consequently, the developed mathematical models for predicting fungal growth on a macroscopic scale can be used as a tool for risk assessment of timber decay in buildings.     

Sensitivity of Nostoc commune UTEX 584 (Cyanobacteria) to water stress

Cells of Nostoc commune UTEX 584 from liquid cultures expressed an upshift in nitrogenase activity when immobilised on inert supports and exposed to matric water potentials between -1.10 and -99.5 MPa. Cells incubated at 0.10 MPa (a_w=c 1.0) maintained increased activity for at least 48 h following immobilization. At water potentials below -23.1 MPa (a_w=0.85), the upshift was transitory. Nitrogenase activity decreased rapidly when immobilised cells were incubated at lower values of _m.Desiccated cells stored at -99.5 MPa (a_w=0.50) underwent an upshift in nitrogenase activity, and in the size of the intracellular ATP pool, when rewetted with either distilled water or liquid MB_o medium (_o =-0.18 MPa). The upshift in nitrogenase activity was chloramphenicol-sensitive and was preceeded by a lag. The duration of the lag depended on the time taken to equilibrate cells to-99.5 MPa, the time desiccated, and the conditions of storage and rewetting. Cells that had no, or very low, nitrogenase activity when rewetted in air, showed a marked stimulation of nitrogenase activity in the presence of 5% v/v CO₂ under both aerobic and anerobic conditions.When rewetted in the presence of 1% w/v glucose (_o =-0.14 MPa), vegetative cells remained intact, but heterocysts underwent autolysis and nitrogenase activity was not detected, even in the presence of 5% v/v CO₂.Abbreviations TTC 2,3,5-triphenyl-2-tetrazolium chloride - _m matric water potential - _o osmotic water potential - a_w water activity     

Produktion von Ochratoxin A und B durch<Emphasis Type="Italic">Penicillium verrucosum</Emphasis> auf Gerste in Abhängigkeit der Parameter Wasseraktivität,Wassergehalt und Temperatur

The ochratoxin A and B (OTA, OTB) production by a toxigenic isolate ofPenicillium verrucosum grown on brewing barley up to six weeks was studied at a storage temperature of 25 °C and different moisture and water activity conditions. Sorption isothermes for barley were prepared at temperatures of 10°C, 15°C and 25°C. OTA was produced after 2 weeks of storage at moisture contents of ≥19%, which is equivalent to water activities (a_w) of 0.83 (adsorptive) and 0.82 (desorptive) at 25 °C. Increased OTA concentrations (5.8-fold and 16.1-fold) were noticed when the moisture contents were adjusted to 20% (a_{w [ads]} 25 °C=0.86) and 21% (a_{w [ads]} [ 25 °C=0.88), respectively. An increase was also shown during storage of 4 and 6 weeks (1.2-fold and 2.4-fold, respectively). Production of OTB was shown to occur at moisture contents ≥18% (a_{w [ads]} 25 °C=0.81). The findings document that OTA and OTB are not produced byP. verrucosum grown on barley stored below 18% moisture content.     

Electric‐induced mortality of newly transformed juvenile fishes in waters of different conductivity

In three experiments, each with three species of newly transformed juvenile fishes, the immediate mortality was determined after electrical exposure to 60 Hz pulsed DC in waters of different conductivity (C_w). With a constant applied power density (D_a; 1·0–4·9 mW cm^?3 depending on species) over a range of C_w(10–1020 μS cm^?1), the results predicted that the highest fish mortality would occur at C_w of 65 μS cm^?1 for bluegill Lepomis macrochirus, 74 μS cm^?1 for largemouth bass Micropterus salmoides and at 140–175 μS cm^?1 for channel catfish Ictalurus punctatus. In experiment 2, the voltage gradient (E) was maintained constant (2·5–8·0 peak V cm^?1 depending on species) over the same range of C_w, and fish mortality increased with current density (J) or D_a, which are directly related to C_w. In experiment 3, fish mortality did not differ when peak E(3 or 8 V cm^?1 depending on species) and mean J(0·09 or 0·24 mA cm^?2 depending on species) were held constant by changing pulse width in waters with different C_w(99, 165 or 495 μS cm^?1). Fish mortality in this experiment was not significantly related to peak or mean transferred power density, and the ‘power transfer theory for electrofishing’ was not useful for predicting electrofishing mortality. Overall, the results of the present study indicated that mortality caused by exposure to electricity can be predicted more accurately with the variables peak E and mean J than with models requiring determination of effective conductivity of the fish.     

The growth limits of a large number of Listeria monocytogenes strains at combinations of stresses show serotype--and niche-specific traits

Aims: The aim of this study was to associate the growth limits of Listeria monocytogenes during exposure to combined stresses with specific serotypes or origins of isolation, and identify potential genetic markers. Methods and Results: The growth of 138 strains was assessed at different temperatures using combinations of low pH, sodium lactate, and high salt concentrations in brain heart infusion broth. None of the strains was able to grow at pH ≤ 4·4, a_w ≤ 0·92, or pH ≤ 5·0 combined with a_w ≤ 0·94. In addition, none of the strains grew at pH ≤ 5·2 and NaLac ≥ 2%. At 30°C, the serotype 4b strains showed the highest tolerance to low pH and high NaCl concentrations at both pH neutral (pH 7·4) and mild acidic conditions (pH 5·5). At 7°C, the serotype 1/2b strains showed the highest tolerance to high NaCl concentrations at both pH 7·4 and 5·5. Serotype 1/2b meat isolates showed the highest tolerance to low pH in the presence of 2% sodium lactate at 7°C. ORF2110 and gadD1T1 were identified as potential biomarkers for phenotypic differences. Conclusions: Differences in growth limits were identified between specific L. monocytogenes strains and serotypes, which could in some cases be associated with specific genetic markers. Significance and Impact of the Study: Our data confirm the growth limits of L. monocytogenes as set out by the European Union for ready-to-eat foods and provides an additional criterion. The association of L. monocytogenes serotypes with certain stress responses might explain the abundance of certain serotypes in retail foods while others are common in clinical cases.     

Two additives to improve stability of immobilized lipase

Abstract

The presence of two different additives during non-covalent immobilization of lipase was studied. Lipase was immobilized via hydrophobic interactions on an amorphous silica with large pore size bearing octyl groups on the surface. Polyethyleneglycol (PEG) with different molecular weights (MW: 1500, 3000 and 10,000) were added to the suspension during enzyme immobilization, in an enzyme to PEG molar ratio of 1:10, and also 1:20 in the case of PEG1500. The activity after 15 d increased from 10% (absence of PEG) to values close to 40% in samples with PEG except the catalyst immobilized in the presence of 1:10 PEG1500, which kept fully active after 15 d incubation in toluene at 70?°C. The presence of water during storage of immobilized enzymes leads to significant activity loss. Saturated solutions of salts controlling the water activity of the systems were used to reduce in a controlled fashion the moisture of the systems: CaCl₂ (a_w=?0.037), MgCl₂ (a_w=?0.328), Mg(NO₃)₂ (a_w=?0.529), Na₂PO₄.12H₂O (a_w=?0.74) and KCl (a_w=?0.84). The immobilized lipase was suspended in saturated solutions of these salts, and then filtered and incubated in desiccators in the presence of the corresponding saturated salt solutions. Catalysts suspended and incubated in KCl or only suspended in phosphate kept some 20% activity after 33 d incubation whereas the maximal stability was achieved when the catalyst was suspended in phosphate and kept in a desiccator without salt solution. This catalyst kept around 50% activity after 33 d incubation. An inversely proportional relationship can be established between the stability achieved by the enzyme and the water content of the system.     

Water activity‐temperature state diagrams of freeze‐dried Lactobacillus acidophilus (La‐5): Influence of physical state on bacterial survival during storage

Water activity‐temperature state diagrams for Lactobacillus acidophilus freeze‐dried in a sucrose or a lactose matrix were established based on determination of stabilized glass transition temperatures by differential scanning calorimetry during equilibration with respect to water activity at fixed temperatures. The bacteria in the lactose matrix had higher stabilized glass transition temperatures for all a_w investigated. The survival of Lactobacillus acidophilus determined as colony forming units for up to 10 weeks of storage at 20°C for (i) a_w = 0.11 with both freeze‐dried matrices in the glassy state, (ii) a_w = 0.23 with the bacteria in the lactose matrix in a glassy state but with the bacteria in sucrose matrix in the nonglassy state, and (iii) a_w = 0.43 with both freeze‐dried matrices in a nonglassy state showed that the nature of the sugar was more important for storage stability than the physical state of the matrix with the nonreducing sucrose providing better stability than the reducing lactose. © 2009 American Institute of Chemical Engineers Biotechnol. Prog., 2009