Microbiome composition and implications for ballast water classification using machine learning

Microbiome Signature of Endophytes in Wheat Seed Response to Wheat Dwarf Bunt Caused by Tilletia controversa Kühn

Wheat dwarf bunt leads to the replacement of seeds with fungal galls containing millions of teliospores of the pathogen Tilletia controversa Kühn. As one of the most devastating internationally quarantined wheat diseases, wheat dwarf bunt spreads to cause distant outbreaks by seeds containing teliospores. In this study, based on a combination of amplicon sequencing and isolation approaches, we analyzed the seed microbiome signatures of endophytes between resistant and susceptible cultivars after infection with T. controversa. Among 310 bacterial species obtained only by amplicon sequencing and 51 species obtained only by isolation, we found 14 overlapping species by both methods; we detected 128 fungal species only by amplicon sequencing, 56 only by isolation, and 5 species by both methods. The results indicated that resistant uninfected cultivars hosted endophytic communities that were much more stable and beneficial to plant health than those in susceptible infected cultivars. The susceptible group showed higher diversity than the resistant group, the infected group showed more diversity than the uninfected group, and the microbial communities in seeds were related to infection or resistance to the pathogen. Some antagonistic microbes significantly suppressed the germination rate of the pathogen's teliospores, providing clues for future studies aimed at developing strategies against wheat dwarf bunt. Collectively, this research advances the understanding of the microbial assembly of wheat seeds upon exposure to fungal pathogen (T. controversa) infection. IMPORTANCE This is the first study on the microbiome signature of endophytes in wheat seed response to wheat dwarf bunt caused by Tilletia controversa Kühn. Some antagonistic microbes suppressed the germination of teliospores of the pathogen significantly, which will provide clues for future studies against wheat dwarf bunt. Collectively, this research first advances the understanding of the microbial assembly of wheat seed upon exposure to the fungal pathogen (T. controversa) infection.

Pathogenic hitchhiker diversity on international ships' ballast water at West Malaysia port

This study investigates bacterial diversity and potential pathogens in the international ships' ballast water at Tanjung Pelepas Port, Malaysia, using 16S rRNA amplicon sequencing. Thirty-four bacterial phylum, 305 families, 577 genera, and 941 species were detected in eight ballast water samples of different origins. The similarity of the bacterial composition between samples was found to be random and not tied to geographical locations. The bacterial abundance did not seem to be affected by related physicochemical except for temperature. Ballast water samples with a temperature lower than 25 °C showed a relatively lower bacterial abundance. A total of 33 potential pathogens were detected from all ballast water samples. Pseudomonas spp., Tenacibaculum spp., Flavobacteriaceae spp., Halomonas spp., and Acinetobacter junii are the potential pathogens with more than 10% OTU prevalence. This study would provide beneficial information for further enhancing ballast water microorganism guidelines in Malaysia.

Machine learning applications in microbial ecology, human microbiome studies, and environmental monitoring

Advances in nucleic acid sequencing technology have enabled expansion of our ability to profile microbial diversity. These large datasets of taxonomic and functional diversity are key to better understanding microbial ecology. Machine learning has proven to be a useful approach for analyzing microbial community data and making predictions about outcomes including human and environmental health. Machine learning applied to microbial community profiles has been used to predict disease states in human health, environmental quality and presence of contamination in the environment, and as trace evidence in forensics. Machine learning has appeal as a powerful tool that can provide deep insights into microbial communities and identify patterns in microbial community data. However, often machine learning models can be used as black boxes to predict a specific outcome, with little understanding of how the models arrived at predictions. Complex machine learning algorithms often may value higher accuracy and performance at the sacrifice of interpretability. In order to leverage machine learning into more translational research related to the microbiome and strengthen our ability to extract meaningful biological information, it is important for models to be interpretable. Here we review current trends in machine learning applications in microbial ecology as well as some of the important challenges and opportunities for more broad application of machine learning to understanding microbial communities.

Metabarcoding quantifies differences in accumulation of ballast water borne biodiversity among three port systems in the United States

Characterizing biodiversity conveyed in ships' ballast water (BW), a global driver of biological invasions, is critically important for understanding risks posed by this key vector and establishing baselines to evaluate changes associated with BW management. Here we employ high throughput sequence (HTS) metabarcoding of the 18S small subunit rRNA to test for and quantify differences in the accumulation of BW-borne biodiversity among three distinct recipient port systems in the United States. These systems were located on three different coasts (Pacific, Gulf, and Atlantic) and chosen to reflect distinct trade patterns and source port biogeography. Extensive sampling of BW tanks (n = 116) allowed detailed exploration of molecular diversity accumulation. Our results indicate that saturation of introduced zooplankton diversity may be achieved quickly, with fewer than 25 tanks needed to achieve 95% of the total extrapolated diversity, if source biogeography is relatively limited. However, as predicted, port systems with much broader source geographies require more extensive sampling to estimate diversity, which continues to accumulate after sampling >100 discharges. The ability to identify BW sources using molecular indicators was also found to depend on the breadth of source biogeography and the extent to which sources had been sampled. These findings have implications both for the effort required to fully understand introduced diversity and for projecting risks associated with future changes to maritime traffic that may increase source biogeography for many recipient ports. Our data also suggest that molecular diversity may not decline significantly with BW age, indicating either that some organisms survive longer than recognized in previous studies or that nucleic acids from dead organisms persist in BW tanks. We present evidence for detection of potentially invasive species in arriving BW but discuss important caveats that preclude strong inferences regarding the presence of living representatives of these species in BW tanks.

Bacterial community composition and diversity in the ballast water of container ships arriving at Yangshan Port, Shanghai, China

Ballast water is a major vector of invasion by protozoans and metazoans. Bacterial invasion is less-well understood. We surveyed the bacterial diversity of ballast water from 26 container ships arriving at the Yangshan Deepwater Port, Shanghai, China during 2015-2016. We characterized the ballast microbiome using high-throughput sequencing (HTS) based on V4-V5 region of 16S rRNA genes. We simultaneously monitored physicochemical parameters of the ballast water, including temperature, pH, dissolved oxygen (DO), salinity, turbidity, total suspended solid (TSS), particulate organic carbon (POC), NO₂, NH₄, PO₄. Proteobacteria was the dominant phylum, comprising more than 50% of the OTUs of almost all vessels, followed by Bacteroidetes (12.08%), Actinobacteria (4.86%) Planctomycetes (3.24%) and Cyanobacteria (1.95%). The relative abundance of Cyanobacteria differed among vessels. It was negatively correlated with temperature, NO₃, pH, TSS, PO₄, and turbidity and positively correlated with NH₄, POC. The genus Synechococcus was the most common Cyanobacteria in our results. Escherichia coli were relatively rare; they are indicator-species of D-2 standards published by the IMO. The relative abundance of the genus Vibrio ranged from 0.003% to 24.88% among different vessels. Our results showed that HTS was able to profile the bacterial communities in ballast-waters, even when the approach was restricted by technical and other obstacles.

Elucidating Biodiversity Shifts in Ballast Water Tanks during a Cross-Latitudinal Transfer: Complementary Insights from Molecular Analyses

In this study, the evolution of ballast water (BW) assemblages across different trophic levels was characterized over a 21 day cross-latitudinal vessel transit using a combination of molecular methods. Triplicate BW samples were collected every second day and size-fractionated (<2.7, 10, >50 μm). Measurements of adenosine triphosphate (ATP) and metabarcoding of environmental nucleic acid (DNA and RNA) analyses, complemented by microscopy and flow cytometry, were performed on each sample. Measured ATP concentrations exhibited high variance between replicates and a strong negative trend in the large (≥50 μm) fraction over the voyage. In concert with microscopy, the metabarcoding data indicated a die-off of larger metazoans during the first week of study and gradual reductions in dinoflagellates and ochrophytes. The ATP and metabarcoding data signaled persistent or increased cellular activity of heterotrophic bacteria and protists in the BW, which was supported by flow cytometry. The metabarcoding showed the presence of active bacteria in all size fractions, suggesting that the sequential filtration approach does not ensure taxonomical differentiation, which has implications for BW quality assessment. Although our data show that ATP and metabarcoding have potential for indicative BW screening for BW compliance monitoring, further research and technological development is needed to improve representativeness of sampling and deliver the unequivocal response criteria required by the international Ballast Water Management Convention.