Gut microbial communities are highly dynamic and respond to factors such as diet and host metabolism. H owever, there is limited knowledge on gut microbiota changes during the daily cycle. In aquaculture, the characterization of gut microbiome dynamics is necessary to identify and understand the relationship it have on host physiology and fitness. To investigate how interactions between feeding and water temperature affected the diel dynamics and activity of the microbiota, we performed two experiments with water temperature and feeding regime as experimental variables, respectively. The qualitative and quantitative composition of the active bacterial community in feces was analyzed along a day cycle in greater amberjack (Seriola dumerili) juveniles.
Materials and methods
In the first experiment , with water temperature as experimental variable, f ish were reared in three independent RAS systems set to 18, 22 and 26°C, under a light/dark photoperiod and fed ad libitum three times daily with a commercial diet. In the second experiment, the fish had two different feeding regimes, continuous feeding or intermittent feeding (three meals per day ), both under continuous illumination. Water temperature was maintained at 22ºC. Sampling was performed along a 24h cycle. For quantitative abundance and activity, we quantified copy numbers of total 16S rDNA genes and rRNA per gram of feces by quantitative PCR (qPCR), and used these data as a proxy of absolute bacterial density and protein synthesis, respectively. RNA:DNA ratio was used as an approximation for the specific activity of bacteria . T he qualitative composition of the active bacterial community was analyzed by r RNA-based amplicon sequencing.
Results from the first experiment identified feeding as the main driver modulating the diel variations in bacterial community activity. A peak in bacterial activity (based on RNA:DNA ratio) was observed at 6 h after first meal for the three temperatures (Fig. 1).
T he composition of the active bacteria community in fish feces varied in relation to postprandial time . At th e order level, feeding promoted a decrease in the relative abundance of active Spirochaetes and an increase in Vibrionales . Although the relative abundance of active Mycoplas moidales was stable throughout the day in all treatments , the absolute abundance of active bacteria peaked at first postprandial sampling (6 h) for this order, being the order mainly responsible for the increase in absolute bacterial activity . At the lowest water temperature, the diel variations were slower.
Results from the second experiment confirmed the role of feeding as a synchronizer of the diel d ynamics of the activity of the bacterial communit ies in the fish gut. F ish fed continuously did not have significant daily variations either in quantitative or qualitative composition of the active bacterial community . By contrast , and in agreement with the data shown above, intermittently fed- fish revealed a peak in specific bacterial activity after the first daily meal and postprandial changes in the relative abundance of active bacteria along the day . At t he order level, t he composition of the active bacterial community in continuously fed fish was dominated by Vibrionales during the whole day cycle, with minor fluctuations in its relative abundance (Fig. 2) .
PERMANOV A based on the Sørensen similarity indicated that communities’ composition from the two feeding regimes, based on the ASV table, were different depending on sampling time. Active bacterial communities were significant different a t the beginning of the day, before the first meal was offered to the intermittently fed-fish ( 0 h, p<0.05 ) and close to being different (p<0.08) at the end of the day cycle (8 and 12 h after the last meal for the intermittent fed-fish).
This is the first study assessing the quantitative and qualitative effects of water temperature and daily feeding rhythm on the gut microbiota dynamics in a poikilothermic animal. 16S rRNA:rDNA ratio revealed a postprandial increase in bacterial-specific activity , pointing to feeding as the main driver of the bacterial activity dynamics . The most noticeable change in the composition of the active bacteria community was a postprandial increase in the order Vibrionales, but for absolute abundance, Mycoplasmoidales was the most responsive order. Thus, the absolute abundance of active bacteria is more suitable for analyzing bacterial dynamics in a community . Low water temperature affected the timeline of the diel pattern in bacterial activity by slowing down the response, probably due to a lower maximum grow th rate of the bacteria and a slower gut transit time in fish.
The composition of active bacteria community in c ontinuously fed fish kept an “active-feeder” profile throughout the whole day, with Vibrionales as the dominant order in terms of both, relative and absolute abundances . Sørensen similarity indicated significant differences in active bacteria community between “pre-feeding” and “active-feeder” fish.