The bacteriocin producer Lactobacillus plantarum LPCO10, a strain originally isolated from an olive fermentation, was used as a starter culture for traditional Spanish-style green olive fermentation. By means of a 2 3 (7-4) fractional factorial design, the salt concentration in brines was found to be the most important factor for producing the highest acidity and the highest initial population of the starter culture. To a lesser extent, other factors such as a high inoculum size, use of MRS as carrier for the starter culture, inoculation time, and type of acid to correct the pH had also an overall significant effect on the final acidity or total corrected acidity. For all initial conditions investigated, Lb. plantarum LPCO10 was able to dominate over the natural population of lactic acid bacteria and leaded to a faster decrease of pH and a faster acidification than the spontaneous process during the first 25 days after brining. Then, wild lactobacilli grew slowly and the population of lactic cocci raised, while LPCO10 decreased. Inoculation also stimulated the growth of yeasts. Thus, the traditional fermentation of Spanish green table olives could be improved by using Lb. plantarum LPCO10 (suspended in MRS) as a starter culture at 10 7 cfu/ml of brine and well defined starting brine parameters i.e.,  4% w/v of NaCl, and initial pH correction (from 4.5 to 6.5) with acetic acid. Inoculation can take place in 1-4 days after brining.

Optimization of bacteriocin production by Lactobacillus plantarum LPCO10 was explored by an integral statistical approach. In a prospective series of experiments, glucose and NaCl concentrations in the culture medium, inoculum size, aeration of the culture, and growth temperature were statistically combined using an experimental 2 3 5-2 fractional factorial two-level design and tested for their influence on maximal bacteriocin production by L. plantarum LPCO10. After the values for the less-influential variables were fixed, NaCl concentration, inoculum size, and temperature were selected to study their optimal relationship for maximal bacteriocin production. This was achieved by a new experimental 3 2 3-1 fractional factorial three-level design which was subsequently used to build response surfaces and analyzed for both linear and quadratic effects. Results obtained indicated that the best conditions for bacteriocin production were shown with temperatures ranging from 22 to 27°C, salt concentration from 2.3 to 2.5%, and L. plantarum LPCO10 inoculum size ranging from 10 7.3 to 10 7.4 CFU/ml, fixing the initial glucose concentration at 2%, with no aeration of the culture. Under these optimal conditions, about 3.2 ؋ 10 4 times more bacteriocin per liter of culture medium was obtained than that used to initially purify plantaricin S from L. plantarum LPCO10 to homogeneity. These results indicated the importance of this study in obtaining maximal production of bacteriocins from L. plantarum LPCO10 so that bacteriocins can be used as preservatives in canned foods.Lactic acid bacteria (LAB) are widely used as starter cultures in dairy, meat, and vegetable fermentations (1, 7, 36). One major reason for their wide use is the wide range of antimicrobial substances that they are able to produce which efficiently contribute to the preservation of the fermented products (13,15,31). Of these antimicrobial substances, bacteriocins are one of the most promising natural food preservatives produced by LAB (10,12,26,33,34). This preservation potential could be achieved either by using a bacteriocin-producing starter culture or by applying the bacteriocin itself as a food additive. The latter will necessarily require optimization of their production, for this is dependent on multiple factors which are usually strain specific (9, 39).In the past, several studies have pursued this goal for a number of different bacteriocins. They have generally focused on the effects of pH, temperature, composition of the culture medium, and general microbial growth conditions (in vitro as well as in natural fermentations) on maximal bacteriocin production (2,4,6,8,11,14,24,25,(28)(29)(30)(38)(39)(40). However, although some of these studies claim validation by a statistical test, usually variance analysis, the combination of variables and their values and limits are arbitrarily chosen, based mainly on subjective personal experience. Thus, usually no previous prospective experimental design is performed to optimize the information that can be gained ...