Millet and tiger nuts were purchased at the local market in the city of Bobo-Dioulasso, Burkina Faso.

Freeze-dried capsules containing lactic bacteria strains from the Trunature and Spring Valley brands were used. These strains consisted of:

The yeast powder (Saccharomyces cerevisiae Boulardii), in the form of ultra-pharmaceutical yeast, was also used.

The cow’s milk was sterilized by Ultra High Temperature (UHT) treatment for the evaluation of bacterial vitality. Culture medium was purchased from Liofilchem for the isolation of microbial strains: Salted tryptone (0.009%, pH = 7.0 ± 0.2), Sabouraud agar (0.005% chloramphenicol and 4% glucose), and MRS agar.

The tiger nut milk produced according to the process described by Oyedele et al. [24] was used as the culture medium for the preparation of two kinds of sourdough. Sourdough 1 was prepared with Bifidobacteria and Lactobacillus (1.2 × 10¹¹ LB CFU/g). Sourdough 2 was made with Bifidobacterium, Lactobacillus, and yeast (1.2 × 10⁹ LB CFU/g + 4.6 × 10⁵ S CFU/g). They were then stored at 4˚C until use.

The millet and tiger nut flours obtained by adapting the processes described by Oyedele et al. [24] were rolled; then the resulting granules were steamed. Cookies enriched with sourdoughs or without (negative control) were produced according to five main steps recorded in Figure 1 .

Fermentation allows a rapid decrease in pH. The effects of technological treatments on the fermentation capacity of probiotics during the process were evaluated through the measurement of pH during fermentation-dehydration (37˚C and 40˚C) according to the AOAC (Volume 1, 15th Edition, 1990) method. The pH was measured every three hours for 15 hours by randomly sampling 10 g from each batch of biscuits. The relative rates of pH decline (ΔpH) and relative humidity decline (ΔRH) were calculated using Equation (1) and Equation (2) below:

$\Delta \text{pH}=\frac{\left(\text{pHtn}-\text{pHtn}-1\right)}{\text{pHtn}-1}\times 100$ (1)

$\Delta \text{RH}=\frac{\left(\text{RHtn}-\text{RHtn}-1\right)}{\text{RHtn}-1}\times 100$ (2)

The post-process vitality of probiotics was evaluated by measuring their lactofermentative capacity within the produced biscuits. A sample of UHT milk (20 mL) was inoculated with biscuit powder (1 g). Another sterile sample of UHT milk (20 mL) without biscuit powder was used as a negative control. The inoculated milk samples were incubated at 37˚C; then the relative rates of pH reduction were monitored for 12 hours on D + 1 and at D + 90 (storage at 30˚C). ΔpH was calculated using Equation (1).

The effects of technological treatments and storage on the survival rate of probiotics were evaluated through microbiological analyses on D + 1 and D + 90 (storage at 30˚C). A random sample of biscuits (10 g) was crushed in sterile physiological water using sterile gloves. A suspension (10 mL) of the crushed biscuits was then taken under sterile conditions, and a series of successive decimal dilutions was performed. The isolation was performed according to ISO 21527-1:2008 [24] . For bacteria, samples were carried out after inoculation of specific selective MRS agar medium, and incubation at 37˚C under CO₂ for 72 hours (bacteria). For yeasts, samples were carried out after inoculation of specific selective Sabouraud chloramphenicol agar, and incubation at 25˚C for 5 to 7 days (yeasts). The number of Colony-Forming Units (CFUs) per gram of product was determined using culture plates from two successive dilutions, at least one of which had a minimum of 15 colonies, using the formula below:

$N=\frac{{\displaystyle \sum C}}{1.1\times d\times V}$(3)

N: Number of CFU/g of the sample.

∑C: Sum of colonies from the two successive plates selected.

d: First dilution selected.

V: Volume of inoculum (1 mL).

The survival rate of microorganisms was evaluated using the following formula:

$\text{Survival}\text{rate}=\frac{\text{Number}\text{of}\text{CFU}\text{by}\text{g}\text{of}\text{biscuits}}{\text{Number}\text{of}\text{CFU}\text{by}\text{g}\text{of}\text{sourdough}}\times 100$ (4)

Three shots were used for each measurement. Excel Office 365 and R Studio 3.14 software were used for data processing and analysis, as well as for designing graphs. Data are reported either as mean ± Standard Deviation (SD) or, when the distribution deviated from normality, as median with the corresponding Interquartile Range (IQR). The Shapiro-Wilk test was applied to verify distribution normality.

Comparisons of proportions between independent groups with small expected frequencies were performed using Fisher’s exact test. For paired categorical data, McNemar’s exact test was applied.

Median differences between two independent groups were assessed using the Mann-Whitney U test, while comparisons involving more than two groups relied on the Kruskal-Wallis test with suitable post hoc procedures when needed. The Wilcoxon signed-rank test was used to compare values in paired measurements. Statistical significance was established at a p-value threshold of <0.05.

The study evaluated the impact of technological treatments on fermentation. Capacity and the vitality of probiotics were determined by measuring the relative rates of pH decrease (ΔpH), as a function of the variation in the relative rates of humidity lowering (ΔHR). Figure 2 shows the effects of technological treatments on the evolution of the pH of biscuits. Globally, three phases of pH evolution were observed during fermentation-dehydration. A first phase (T0 - T6) of rapid deceleration of the pH-lowering rates occurred concomitantly with an acceleration of

the relative humidity lowering rates. The highest pH reduction rates were observed successively with LBS biscuits at 40˚C and 37˚C; followed by LB biscuits at 37˚C and 40˚C. The lowest rates were recorded with the control biscuits. The second phase (T6 - T9) of the evolution of ΔpH was characterized by a stabilization of pH and a peak in the reduction of humidity. The third phase (T9 - T15) of evolution was marked by a slight increase in pH reduction rates and a progressive regression in the lowering of hygrometry.

The determination of lactofermentative (UHT milk) capacities of bacteria allowed the evaluation of the vitality of probiotics. Three-hourly pH monitoring showed a continuous increase in pH reduction rates on Day + 1 and Day + 90 ( Figure 3(a) and Figure 3(b) ). The highest reduction rates were observed successively with biscuits treated at 37˚C (with LB and LBS) and at 40˚C (with LB and LBS). The best pH reduction rates were noted at Day + 1.

Table 1 describes the survival rates of probiotics inoculated into biscuits after technological processing and storage. The best survival rates were obtained with processing at 37˚C on Day + 1 and during storage (Day + 90). The highest survival rates were observed at Day + 1 with yeast multiplication (313.04%). Lactic bacteria in co-culture with yeast had a higher survival rate (32.50%) than lactic bacteria in monoculture (1.83%). A significant reduction in survival rates was recorded during storage.

D + 1 = 1 day after biscuit production; D + 90 = 90 days after biscuit production.