Jean-Baptiste Aubin

Human-Driven Microbiological Contamination of Benthic and Hyporheic Sediments of an Intermittent Peri-Urban River Assessed from MST and 16S rRNA Genetic Structure Analyses

Rivers are often challenged by fecal contaminations. The barrier effect of sediments against fecal bacteria was investigated through the use of a microbial source tracking (MST) toolbox, and by Next Generation Sequencing (NGS) of V5-V6 16S rRNA gene (rrs) sequences. Non-metric multi-dimensional scaling analysis of V5-V6 16S rRNA gene sequences differentiated bacteriomes according to their compartment of origin i.e., surface water against benthic and hyporheic sediments. Classification of these reads showed the most prevalent operating taxonomic units (OTU) to be allocated to Flavobacterium and Aquabacterium. Relative numbers of Gaiella, Haliangium, and Thermoleophilum OTU matched the observed differentiation of bacteriomes according to river compartments. OTU patterns were found impacted by combined sewer overflows (CSO) through an observed increase in diversity from the sewer to the hyporheic sediments. These changes appeared driven by direct transfers of bacterial contaminants from wastewaters but also by organic inputs favoring previously undetectable bacterial groups among sediments. These NGS datasets appeared more sensitive at tracking community changes than MST markers. The human-specific MST marker HF183 was strictly detected among CSO-impacted surface waters and not river bed sediments. The ruminant-specific DNA marker was more broadly distributed but intense bovine pollution was required to detect transfers from surface water to benthic and hyporheic sediments. Some OTU showed distribution patterns in line with these MST datasets such as those allocated to the Aeromonas, Acinetobacter, and Pseudomonas. Fecal indicators (Escherichia coli and total thermotolerant coliforms) were detected all over the river course but their concentrations were not correlated with MST ones. Overall, MST and NGS datasets suggested a poor colonization of river sediments by bovine and sewer bacterial contaminants. No environmental outbreak of these bacterial contaminants was detected.

Bacteriome genetic structures of urban deposits are indicative of their origin and impacted by chemical pollutants

Urban activities generate surface deposits over impervious surfaces that can represent ecological and health hazards. Bacteriome genetic structures of deposits washed off during rainfall events, over an urban industrial watershed, were inferred from 16 S rRNA gene (rrs) sequences generated by high throughput sequencing. Deposits were sampled over a 4 year-period from a detention basin (DB). Major shifts, matching key management practices, in the structure of these urban bacteriomes, were recorded. Correlation analyses of rrs similarities between samples and their respective concentrations in chemical pollutants, markers of human fecal contaminations (HF183) and antimicrobial resistances (integrons), were performed. Harsher environmental constraints building up in the older deposits led to an increase number of rrs reads from extremophiles such as Acidibacter and Haliangium. Deposits accumulating in the decantation pit of the DB showed an increase in rrs reads from warm blooded intestinal tract bacteria such as Bacteroides and Prevotella. This enrichment matched higher concentrations of Bacteroides HF183 genotypes normally restricted to humans. Bacteriomes of urban deposits appeared good indicators of human-driven environmental changes. Their composition was found representative of their origin. Soil particles and rain appeared to be major contributors of the inferred bacterial taxa recovered from recent deposits.

Projection density estimation under a m-sample semiparametric model

An m-sample semiparametric model in which the ratio of m-1 probability density functions with respect to the mth is of a known parametric form without reference to any parametric model is considered. This model arises naturally from retrospective studies and multinomial logistic regression model. A projection density estimator is constructed by smoothing the increments of the maximum semiparametric empirical likelihood estimator of the underlying distribution function, using the combined data from all the samples. Some asymptotic results on the proposed projection density estimator are established. Connections between our estimator and kernel semiparametric density estimator are pointed out. Some results from simulations and from the analysis of two real data sets are presented.

An asymptotic factorization of the Small–Ball Probability: theory and estimates

This work reviews recent results on an asymptotic factorization of the Small–Ball Probability of a \( {\fancyscript{L}}_{[0,1]}^{2} \)–valued process, as the radius of the ball tends to zero. This factorization involves a volumetric term, a pseudo–density for the probability law of the process, and a correction factor. Estimators of the latter two factors are introduced and some of their theoretical properties considered.