Gut microbiota of invasive Asian silver carp (SVCP) and indigenous planktivorous gizzard shad (GZSD) in Mississippi river basin were compared using 16S rRNA gene pyrosequencing. Analysis of more than 440 000 quality-filtered sequences obtained from the foregut and hindgut of GZSD and SVCP revealed high microbial diversity in these samples. GZSD hindgut (GZSD_H) samples (n=23) with >7000 operational taxonomy units (OTUs) exhibited the highest alpha-diversity indices followed by SVCP foregut (n=15), GZSD foregut (n=9) and SVCP hindgut (SVCP_H) (n=24). UniFrac distance-based non-metric multidimensional scaling (NMDS) analysis showed that the microbiota of GZSD_H and SVCP_H were clearly separated into two clusters: samples in the GZSD cluster were observed to vary by sampling location and samples in the SVCP cluster by sampling date. NMDS further revealed distinct microbial community between foregut to hindgut for individual GZSD and SVCP. Cyanobacteria, Proteobacteria, Actinobacteria and Bacteroidetes were detected as the predominant phyla regardless of fish or gut type. The high abundance of Cyanobacteria observed was possibly supported by their role as the fish’s major food source. Furthermore, unique and shared OTUs and OTUs in each gut type were identified, three OTUs from the order Bacteroidales, the genus Bacillariophyta and the genus Clostridium were found significantly more abundant in GZSD_H (14.9–22.8%) than in SVCP_H (0.13–4.1%) samples. These differences were presumably caused by the differences in the type of food sources including bacteria ingested, the gut morphology and digestion, and the physiological behavior between GZSD and SVCP.
The prevention of non-indigenous aquatic invasive species spreading into new areas is a goal of many resource managers. New techniques have been developed to survey for species that are difficult to capture with conventional gears that involve the detection of their DNA in water samples (eDNA). This technique is currently used to track the invasion of bigheaded carps (silver carp and bighead carp; Hypophthalmichthys molitrix and H. nobilis) in the Chicago Area Waterway System and Upper Mississippi River. In both systems DNA has been detected from silver carp without the capture of a live fish, which has led to some uncertainty about the source of the DNA. The potential contribution to eDNA by vectors and fomites has not been explored. Because barges move from areas with a high abundance of bigheaded carps to areas monitored for the potential presence of silver carp, we used juvenile silver carp to simulate the barge transport of dead bigheaded carp carcasses, slime residue, and predator feces to determine the potential of these sources to supply DNA to uninhabited waters where it could be detected and misinterpreted as indicative of the presence of live bigheaded carp. Our results indicate that all three vectors are feasible sources of detectable eDNA for at least one month after their deposition. This suggests that current monitoring programs must consider alternative vectors of DNA in the environment and consider alternative strategies to minimize the detection of DNA not directly released from live bigheaded carps.