Perna is an edible novel cook-then-lactic ferment product made from mussels, particularly the New Zealand green shell mussel, Perna canaliculus. This study characterized Perna form a physicochemical and microbiological perspective. Perna is made in the sequence: cook mussels by boiling, shuck, comminute, add glucose, salt, a carrageenan thickener and a starter culture; evacuate in barrier bags and incubate at 30°C for 4 days; store as required. The aim of this thesis was to confirm the potential to develop a safe and stable product. The input variables included annual mussel condition, mussel gender, post fermentation storage conditions and starter culture type. Most work was done with a single lactic culture. Output variables include colour, pH, soluble protein, soluble peptides, amino nitrogen, profiles of free fatty acids and free amino acids, and of volatile compounds. Conventional microbiological plating techniques revealed the microbiological profile at stages of production and storage, and the capability of different starters to inhibit Listeria monocytogenes and, indirectly, Clostridium botulinum. In the study of Perna’s characteristics due to annual variation in mussel quality, Perna were prepared over many months and monitored for colour, pH, gender proportion and basic microbiology. The mussels used were as bought at retail with no supply chain history other than date. The pH of Perna at day of preparation had no obvious pattern with season. The pH changes at completion of showed no obvious pattern. On subsequent storage at 4°C Perna pH continued to decline slowly with storage. The orange/apricot colour for female and creamy white for male gonads was largely swamped by the brown colour of viscera after comminution. Long storage at ambient (19-23 ° C) temperature markedly darkened Perna, but colour was stable at chill temperature. Thus, storage at 4°C is best option for Perna stability, but the temperature requirements are not particularly rigorous. Limited microbiological showed that during the early phase of fermentation, contaminating microorganisms compete for nutrients with desirable bacteria in starter cultures. Mussels are filter feeders and are known to be highly contaminated with marine microorganisms. Therefore, the cooking step is very important to reduce the native biota. The changes in protein characteristics from surviving endogenous enzymes of mussels and microflora, and of culture bacteria were investigated in aqueous extracts. Titratable amino nitrogen, a measure of chemically available amino groups, showed little change on fermentation, but storage at higher temperature, 35°C, resulted in big increases in amino nitrogen suggesting enzymatic hydrolysis to the point of spoilage. Long term storage at lower temperature resulted in slightly reduced amino nitrogen perhaps arising from the Maillard reaction or loss due to microorganisms converting free amino groups to stable products. The similarity of ultraviolet absorbance between 275 to 292 nm and 200 to 400 nm indicated that proteinaceous matter was dominant in the water-soluble portion of Perna rather than nucleic acids. A loss in soluble protein by the bicinchoninic assay with storage time was clear but could not be easily reconciled with the amino nitrogen results. One plausible reason was proposed. In free amino acid analysis individual amino acids showed seasonal changes, with some amino acids showing an inverse correlation. Proportions of arginine were particularly striking, revealing three amino acid patterns related to season, and attributable to the role of arginine as an energy reserve when mussel food was not limiting. Alanine, glutamic acid and glycine, which are very flavour active, were present in high concentrations in Perna extract. Amino acid analysis work extended to differences due to five Chr Hansen commercial cultures. Two were standouts in respect of acidification rate and changes in amino profile. However, after extended storage amino acid profiles were similar for all five, suggesting a similar ultimate flavour from amino acids (but not necessarily from volatiles). Cultures 4 and 5 were the fastest metabolizers to create an acidic environment in Perna. Lactic acid bacteria and staphylococci contributed to biogenic amine formation in Perna when the fermentation process was delayed, but not with advanced storage where biogenic amines were completely absent. Even at early times, biogenic amine concentrations were below regulated food safety limits. The fatty acid profile of Perna was dominated by palmitic acid, followed by docosahexaenoic acid, then eicosapentaenoic acid, and subsequently a host of more minor fatty acids. The fatty acid profile was tracked throughout Perna preparation and long-term storage. The profile was extraordinarily stable. Thus, cooking, an anaerobic fermentation step and subsequent storage under vacuum is an excellent way of maintaining fatty acid stability. The effect of cultures (five), storage time and temperature, and glucose concentration on Perna volatiles was explored by solid phase microextraction of the headspace above Perna. Different cultures produced different headspace profiles. Storage time and temperature also resulted in differences, conforming that chill storage was best. Below 1.0% glucose, the volatile pattern suggested flavour deterioration, conformed by smell and colour changes. With insufficient glucose it is likely that the surviving microflora successfully competes with the desirable culture microflora. Cooked mussel is not the intended matrix for meat salami cultures; microbial responses are unknown beyond the fact that pH usually fell to preservation levels, provided enough glucose was present.

The microbiological characteristics of Perna prepared using one culture stored at 4°C, ambient and 35°C were studied. Also, determination of the basic microbiology of Perna prepared using five starter cultures was done. Several pathogens – Staphylococcus aureus, Salmonella spp., Escherichia coli, and Vibrio parahaemolyticus – were not detectable in three preparation trials. The ability of five starter cultures inhibit Listeria monocytogenes and Clostridium botulinum was examined. In vitro experiment showed certain strain of Lactobacillus sakei had a marked inhibitory ability on L. monocytogenes. This was in the culture designated T-SC-150. Some lactic acid bacteria have been shown to cause a spontaneous mutation on L. monocytogenes but T-SC-150 simply resulted in complete suppression of L. monocytogenes. In situ experiments with supernatant of fermented Perna showed to provide favourable condition for L. monocytogenes inhibition. In this experiment T-SC-150 was able to reduce pH faster than other cultures but all cultures were able to drop pH below 5.2 within 17 hours of fermentation, greatly faster than the requirement of 48 hours by Ministry of Primary Industry. T-SC-150 was the standout in the cooked mussel matrix. Its use as starter culture for Perna seems obvious, particularly as other quality indicators (pH, free amino acids, fatty acids, and volatiles) did not deviate to any extent from averages. There is one important merchandising advantage for Perna. Provided it is stored cool, with some temperature flexibility, the shelf life would be long with a ‘best by’ date months after preparation. Future work is likely to center on recipe development of several Perna formats (very finely comminuted, Perna as described here, whole mussel format), but which was beyond the scope of this thesis. That is work for creative chefs.