According to the recommendations of Boyd (1988) for farming crustaceans, the suitable temperature is in the range of 25–32oC, the suitable pH for good growth is between 6–9, and the reasonable alkalinity should be maintained above 80 mg/L. In this experiment, temperature, pH, and alkalinity were found to be in the range of values mentioned above, so it is considered suitable for the growth of white leg shrimp. Besides, dissolved oxygen (DO) is also considered appropriate for shrimp growth, according to the recommendation of Brock and Main (1994). TAN decreased in all treatments, but the treatments with added rice flour and molasses were lower than the control (p < 0.05). The Deng et al. (2018) study also showed a lower TAN in the additional carbohydrate treatments than in control after day 28th. According to the authors, the improvement in TAN was attributed to ammonia assimilation by heterotrophic bacteria, which immobilized TAN to synthesize new bacteria cells. Besides, treatments with a high ratio of molasses (R50-M50 and R10-M90) have a faster level of TAN removal, which was also mentioned in the study of Wasielesky et al. (2006) that ammonia removal is slower due to the long time it takes to decomposite complex carbohydrates into simple sugars. The treatment R70-M30 had the lowest NO2− concentration than other treatments, indicating less toxicity to shrimp. NO3− concentration showed an almost similar fluctuation tendency to the control, which peaked on day 7th and gradually decreased on day 21st. The nitrate fluctuations during the experiment period ranged from 5.4 to 14.7 mg/L and were within the appropriate range for the Pacific white shrimp growth, according to Azim and Little (2008). The reduction of nitrite and nitrate in this current experiment showed the effectiveness of carbohydrate supplementation in improving water quality in shrimp ponds, as demonstrated by the research of Al-Gahwari and Al-Buhaishi (2021).
TSS and VSS results showed that the fluctuations were within a suitable range for Pacific white shrimp stocked by the biofloc system, as reported by Wasielesky et al. (2013). Figure 2A shows that TSS in control generally fluctuated lower than in the treatments with the addition of carbohydrates. This trend may be due to more floc formation in the carbohydrate treatments. The VSS content in control tended to be lower than the additional carbohydrate treatments (Fig. 2B), indicating that the residual solids in the water (mostly organic matter) were high. The amount of organic matter was contributed by aquatic animal excretion, feed excess, and microbial density, where the content of animal excretion and feed excess may be higher. Rittmann and McCarty (2001) stated that VSS content was a measure of bacterial biomass in water. Floc length increased with the addition of carbohydrates, in which the ratio of rice flour and molasses 70:30 showed better results (significantly different from other treatments except for R50-M50). Floc width also tended to be smaller when no carbohydrate was added. These changes were considered consistent with the report of Duy and Khanh (2018) mentioned that the addition of carbohydrates promoted a change in floc size. Flocs are formed from the adhesion of microorganisms and organic matter in water. Each floc is held by the secretions of bacteria, by bound of filamentous bacteria, or by electrostatic attraction (Hargreaves 2013). The more abundant these components are the greater aggregation of microorganisms, leading to an increase in floc size. According to the study of Duy and Khanh (2018), the accumulation of PHB (poly-B-hydroxybutyrate) contributed to the increase in floc size. Floc volume also recorded an enhancement in volume when the system was supplemented with two carbohydrate sources. Treatments with a rice-flour ratio of 30% gave positive results. This tendency is also demonstrated by the studies of Duy and Khanh (2018) and can be explained through the evolution of microorganism density when protozoa and phytoplankton are presented with higher density in the carbohydrate treatments.
The genus Vibrio is a common pathogen in shrimp. Its antagonist – lactic acid bacteria (LAB), is often added to aquaculture systems to inhibit the growth of this pathogenic bacteria (Thompson et al. 2022). In this present study, the number of vibrios in biofloc was 10 to 15 times higher than in water. However, this value did not exceed 103 CFU/mL if it was calculated by the total density of vibrios in one mL of water, including biofloc, which means the density of vibrios was not enough to cause damage to cultured shrimp. The lower vibrio counts in R70-M30 could be explained by competition for space and nutrients with other microbial communities and limited cell-to-cell communication by the biofloc (Anand et al. 2013; Santhana Kumar et al. 2018), especially when carbohydrates were added. Lactic acid bacteria are said to kill or inhibit pathogenic bacteria's growth through metabolic products such as bacteriocins, bacteriocin-like factors, lactic acid, and hydrogen peroxide (Corr et al. 2007). LAB concentration in biofloc in treatments containing more than 50% rice flour may be due to the complicated structure in the molecular structure of rice flour. For example, rice starch's polymeric and branched structure provided a more excellent binding of biofloc, which was the place for microbial retention. The results of mean floc length confirmed this hypothesis that with rice flour ratios of 50% and 70%, LAB density was higher and lower for the ratio of 30% and 10%. Besides, the R70-M30 treatment demonstrated more clearly the vibrio-inhibitory effect of LAB, in which the presence of a large amount of LAB reduced the concentration of vibrio bacteria in the water.
The higher density of phytoplankton in control can be explained by the lack of carbohydrates. Biofloc was formed less with a smaller size, resulting in clearer water, allowing light to penetrate the water easily, leading to a higher density of algae than the rest of the treatments. This result was also similar to the study of Tinh et al. (2021), showing that the chlorophyll content (indicating the density of algae) was highest in the treatment without added corn starch during the weeks of sampling. Regarding the number of zooplankton, the treatments with the addition of carbohydrates made the zooplankton density higher than the control treatment (p < 0.05). During the process, it was noted that protozoa had a large number at the beginning and decreased at the end of the experiment, in contrast to rotifer because protozoa were consumed by rotifer. Thus, treatment without carbohydrate addition had a high density of phytoplankton, a low density of zooplankton, and opposite results for the additional carbohydrate treatments. This difference could explain why the mean DO concentration was not different among all treatments, but the tendency of variation was proportional to the density of organisms in the water (Kungvankij et al. 1986).
The results of the final average weight showed relatively uniform growth when there was no significant difference between treatments. The impact of adding carbohydrates on shrimp was expressed in terms of final biomass and survival. Final biomass tended to be improved when carbohydrate was added (except R90-M10 was the lowest value and significantly different from R70-M30). Besides, under zero-water exchange conditions, the additional carbohydrate treatments did not show a difference in survival compared to the control, resulting in a similar variation of final biomass. Although the parameters did not show a clear effect, it was demonstrated that there were no negative impacts on shrimp growth when the two carbohydrate sources were combined.