Analysis of dissolved oxygen influencing factors and concentration prediction using input variable selection technique: A hybrid machine learning approach.

Accurate quantification of dissolved oxygen (DO) is critically important for the protection and management of aquatic ecosystems. Successful applications have utilized mechanistic and data-driven models to simulate DO content in aquatic ecosystems. However, mechanistic models present challenges due to their complex and difficult-to-solve conditions, making them less portable. Additionally, data-driven model predictions are hindered by the challenge of numerous input variables, impacting both the running speed and prediction performance of the model. To address these challenges, water quality data and meteorological data of the Tanjiang River were obtained. The maximum information coefficient (MIC) input variable selection technique was employed to identify primary environmental factors influencing DO changes. Furthermore, coupled with support vector regression (SVR), two models (SVR and MIC-SVR) were employed to estimate the DO concentration of the Tanjiang River, and the optimal model was established. The results indicated a shift in the primary pollution factor from ammonia nitrogen to total phosphorus after recent treatment in the Tanjiang River. In comparison with the SVR model, the root mean square error (RMSE) of the MIC-SVR model was reduced by 4.46%, and the Nash-efficiency coefficient (NSE) was improved by 45.85%. In addition, study of kernel function selection revealed that considering as many kernel functions as possible is necessary for improving the performance of the SVR model. Conclusively, the proposed MIC-SVR model serves as an effective tool to analyze the relationship between DO and environmental factors, identifying the primary causes of low DO, and accurately predict the DO concentration in the Tanjiang River (especially in its middle and lower reaches), thus providing a reference for governmental decision-making on water environmental protection and water resource management.

Isolation, identification and whole-genome analysis of an Achromobacter strain with a novel sulfamethazine resistance gene and sulfamethazine degradation gene cluster.

A sulfamethazine (SM2) degrading strain, Achromobacter mucicolens JD417, was isolated from sulfonamide-contaminated sludge using gradient acclimation. Optimal SM2 degradation conditions were pH 7, 36 °C, and 5 % inoculum, achieving a theoretical maximum degradation rate of 48 % at 50 ppm SM2. Cell growth followed the Haldane equation across different SM2 concentrations. Whole-genome sequencing of the strain revealed novel functional annotations, including a sulfonamide resistance gene (sul4) encoding dihydropteroate synthase, two flavin-dependent monooxygenase genes (sadA and sadB) crucial for SM2 degradation, and unique genomic islands related to metabolism, pathogenicity, and resistance. Comparative genomics analysis showed good collinearity and homology with other Achromobacter species exhibiting organics resistance or degradation capabilities. This study reveals the novel molecular resistance and degradation mechanisms and genetic evolution of an SM2-degrading strain, providing insights into the bioremediation of sulfonamide-contaminated environments.

Characteristics and fractionations of sediment oxygen demand in a complex tidal river network area.

Tidal river networks are affected by the tide and influenced by complex factors related to sediment oxygen demand (SOD). In this study, we used chemical inhibition to measure the oxygen consumption of different types of SOD to explore the specific oxygen consumption mechanism of sediments. Then, we evaluated the diffusion fluxes of the sediment-water interface and factors affecting SOD using diffusive gradients in thin films. Total SOD in the tidal river network area of the Pearl River basin was ∼0.5928 g/m2/day, which was 8.47% higher than that in the non-tidal river network area but lower than that in black and odorous water reported previously. In the tidal river network area, biological SOD was 15.6% higher in summer than in winter, and the difference in total SOD was greatly influenced by human activity. We observed a significant effect of sediment on SOD in winter, whereas there were no significant correlations between sediment and SOD in summer. Different particle-size distributions lead to different organic matter contents, resulting in different biological SOD ratios between seasons. Our study found that seasonal tidal changes can affect ion exchange at the sediment water interface, leading to changes in SOD.These findings will be of great significance for the study of phenomena associated with low dissolved oxygen in tidal river networks and provide directions for future sediment pollution control.

Environmental DNA metabarcoding reveals the effect of environmental selection on phytoplankton community structure along a subtropical river.

The Dongjiang River, a major tributary of the Pearl River system that supplies water to more than 40 million people in Guangdong Province and neighboring regions of China, harbors rich biodiversity, including many endemic and endangered species. However, human activities such as urbanization, agriculture, and industrialization have posed serious threats to its water quality and biodiversity. To assess the status and drivers of phytoplankton diversity, which is a key indicator of aquatic ecosystem health, this study used Environmental DNA (eDNA) metabarcoding combined with machine learning methods to explore spatial variations in the composition and structure of phytoplankton communities along the Dongjiang River, including its estuary. The results showed that phytoplankton diversity exhibited spatial distribution patterns, with higher community structure similarity and lower network complexity in the upstream than in the downstream regions. Environmental selection was the main mechanism shaping phytoplankton community composition, with natural factors driving the dominance of Pyrrophyta, Ochrophyta, and Cryptophyta in the upstream regions and estuaries. In contrast, the downstream regions was influenced by high concentrations of pollutants, resulting in increased abundance of Cryptophyta. The random forest model identified temperature, dissolved oxygen, chlorophyll a, NO2-, and NH4+ as the main factors influencing the primary phytoplankton communities and could be used to predict changes during wet periods. This study provides valuable insights into the factors influencing phytoplankton diversity and community composition in the Dongjiang River, and demonstrates the application value of eDNA metabarcoding technique in large-scale, long-distance river biodiversity monitoring.

Co-occurrence of organophosphate esters and phosphorus fractions in river sediments: Implications for pollution prediction and environment risk assessment.

Organophosphate esters (OPEs) and phosphorus (P) are widespread pollutants in aquatic ecosystems, presenting potential ecological risks. However, there is still a lack of comprehensive understanding of their relationships in sediments. In this study, we investigated the co-occurrence and behaviors of the OPEs and P in urban river sediments. The results indicated serious OPE and P pollution in the study area, with substantial spatial variations in the contents and compositions. The OPE congeners and P fractions exhibited different correlations, particularly more significant linear relationships (R = 0.455 - 0.816, p < 0.05) were observed between the aryl-OPEs and P fractions, potentially due to the influence from sources, physicochemical properties, and total organic carbon. About 56 to 71% of variability in predicting the concentrations of aryl-OPE can be explained by the multiple linear regression model using the Fe/Al- and Ca-bound P contents. The study regions exhibited greater aryl-OPEs ecological risks were consistent with the regions with more serious Total P pollution levels. This study represents the first report demonstrating the potential of Fe/Al-P and Ca-P contents in predicting aryl-OPE contents in heavily polluted sediments, providing a useful reference to comprehensively assess the occurrence and environmental behaviors of aryl-OPEs in anthropogenic polluted sediments.

Profiles and spatial distributions of heavy metals, microbial communities, and metal resistance genes in sediments from an urban river.

The occurrence and propagation of resistance genes due to exposure to heavy metals (HMs) in rivers is an emerging environmental issue. Little is known about resistance genes in microbial communities in river sediments with low HM concentrations. The profiles and spatial distributions of HMs, the microbial community, and metal resistance genes (MRGs) were analyzed in sediment samples from the Zhilong River basin in Yangjiang city, near the Pearl River Delta. Concentrations of copper (Cu), cadmium (Cd), lead (Pb), chromium (Cr), and nickel (Ni) were relatively low compared with those in other urban river sediments in China. HM chemical composition and fractions and the structure of the microbial community varied along the main channel, but the composition and abundance of MRGs were relatively homogeneous. Variations in HMs and microbial communities in mid- to upstream areas were related to the presence of tributaries, whose inputs were one of the major factors affecting HM chemical fractions and genera structure in mainstream sediments. There were no significant correlations (p < 0.05) between HM concentrations, bacterial communities, and the MRG profiles; thus, HM concentrations were not the main factor affecting MRGs in sediments. These results contribute to understanding the propagation of MRGs in urban rivers in developing cities.

Concentration estimation of dissolved oxygen in Pearl River Basin using input variable selection and machine learning techniques.

Dissolved oxygen (DO) concentration is an essential index for water environment assessment. Here, we present a modeling approach to estimate DO concentrations using input variable selection and data-driven models. Specifically, the input variable selection technique, the maximal information coefficient (MIC), was used to identify and screen the primary environmental factors driving variation in DO. The data-driven model, support vector regression (SVR), was then used to construct a robust model to estimate DO concentration. The approach was illustrated through a case study of the Pearl River Basin in China. We show that the MIC technique can effectively screen major local environmental factors affecting DO concentrations. MIC value tended to stabilize when the sample size >3000 and EC had the highest score with an MIC >0.3 at both of the stations. The variable-reduced datasets improved the performance of the SVR model by a reduction of 28.65% in RMSE, and increase of 22.16%, 56.27% in R2, NSE, respectively, relative to complete candidate sets. The MIC-SVR model constructed at the tidal river network performed better than nontidal river network by a reduction of approximately 63.01% in RMSE, an increase of 62.36% in NSE, and R2 >0.9. Overall, the proposed technique was able to handle nonlinearity among environmental factors and accurately estimate DO concentrations in tidal river network regions.

[Analysis on nitrogen and phosphorus loading of non-point sources in Shiqiao river watershed based on L-THIA model].

Based on the Long-term Hydrological Impact Assessment (L-THIA) model, the effect of land use and rainfall change on nitrogen and phosphorus loading of non-point sources in Shiqiao river watershed was analyzed. The parameters in L-THIA model were revised according to the data recorded in the scene of runoff plots, which were set up in the watershed. The results showed that the distribution of areas with high pollution load was mainly concentrated in agricultural land and urban land. Agricultural land was the biggest contributor to nitrogen and phosphorus load. From 1995 to 2010, the load of major pollutants, namely TN and TP, showed an obviously increasing trend with increase rates of 17.91% and 25.30%, respectively. With the urbanization in the watershed, urban land increased rapidly and its area proportion reached 43.94%. The contribution of urban land to nitrogen and phosphorus load was over 40% in 2010. This was the main reason why pollution load still increased obviously while the agricultural land decreased greatly in the past 15 years. The rainfall occurred in the watershed was mainly concentrated in the flood season, so the nitrogen and phosphorus load of the flood season was far higher than that of the non-flood season and the proportion accounting for the whole year was over 85%. Pearson regression analysis between pollution load and the frequency of different patterns of rainfall demonstrated that rainfall exceeding 20 mm in a day was the main rainfall type causing non-point source pollution.