Data-driven intelligent identification technologies for hazardous waste (medical waste) were systematically reviewed. General principles for modeling based on artificial intelligence and spectroscopy were summarized. Key challenges, including data acquisition, class imbalance, model generalization, and interpretability, were analyzed. Future research directions were proposed, including AI-assisted spectroscopic modeling, transfer learning, multimodal data fusion, and machine-learning-based modeling of relationships among spectra, structures, and properties. A systematic reference for the intelligent identification and management of hazardous (medical) waste was provided, and guidance to accelerate practical deployment in treatment and disposal, resource recovery, and intelligent monitoring was offered.

The synergistic treatment of municipal sludge with black soldier fly larvae (BSFL) and its effects on nitrogen transformation were investigated in this study. Sludge-based substrates with 0%, 10%, and 30% wheat bran (T2~T4) were used for BSFL treatment, with T1serving as the BSFL-free control. BSFL performance, bioconversion efficiency, and nitrogen transformation characteristics during the sludge treatment were systematically evaluated. Results indicated that the combined application of BSFL and wheat bran significantly enhanced biotransformation efficiency (20.13%~35.24%), alongside increases in crude fat (19.69%~ 26.77%) and protein content (16.26%~24.26%). The best growth performance and nutrient accumulation were observed in the treatment group (T4) with a 30% addition of wheat bran. Moreover, a notable reduction in substrate moisture (7.49%~7.65%), organic matter (16.33%~28.50%), and dissolved organic carbon (DOC) (25.53%~47.06%) were achieved, while pH (8.42~8.63) and electrical conductivity (635~1110μS/cm) were increased by BSFL treatment. Analysis of nitrogen fractions revealed that substrate organic nitrogen degradation was significantly accelerated by BSFL, thereby promoting the ammonification process. A 36.45%~51.09% reduction in total nitrogen (TN) was measured in the post-treatment sludge, accompanied by a marked increase in NH₄⁺-N, and consistently low NO₃^--N levels. Redundancy analysis (RDA) revealed that urease activity was positively correlated with NH₄⁺-N, EC, and pH, suggesting that urease and dehydrogenase activities were regulated by BSFL to facilitate the conversion of organic to inorganic nitrogen forms. Furthermore, protease activity exhibited a close relationship with organic matter content, gradually diminishing as substrate availability decreased. These findings demonstrate that BSFL enhance carbon-nitrogen metabolism by modulating key hydrolytic enzymes, thereby optimizing nitrogen transformation pathways in sludge.

Surface and subsurface soils collected from a bamboo forest at the footslope of an upland were used to explore the effects of pig manure dissolved organic (PM DOM) and its hydrophilic (HI) and hydrophobic (HO) fractions on the adsorption, transport, and retention of doxycycline, enrofloxacin, florfenicol, and sulfamethazine. Total DOM was extracted from pig manure and further separated into HI and HO fractions using XAD-8resin column. According to the results of batch equilibrium experiments, the addition of PM DOM led to varying decreases in the adsorption of enrofloxacin, florfenicol, and sulfamethazine but marked increases in the adsorption of doxycycline; the addition of PM DOM's HIS fraction substantially weakened the adsorption of enrofloxacin and florfenicol by the soils; the addition of PM DOM's HO fraction significantly enhanced the adsorption of doxycycline but markedly reduced the adsorption of florfenicol by the soils. According to the results of saturated transport experiments in repacked soil columns, doxycycline and enrofloxacin could make a breakthrough; the HI fraction of PM DOM played a predominant role in reducing the recovery rate of sulfamethazine in the surface and subsurface soils (by 13.6% and 5.4%, respectively), while both the HI and HO fractions of PM DOM increased the recovery rate of florfenicol (only evidenced in the surface soil; by 30.7% and 20.1%, respectively). The addition of PM DOM could enhance the retention of sulfamethazine and florfenicol (by 14.9% and 7.8%, respectively, in the surface soil columns; by 3.1% and 10.6%, respectively, in the subsurface soil columns) while promote the redistribution of enrofloxacin and doxycycline to deeper soil layers. The humification degree and aromaticity of DOM in the aqueous phase of the soils, as well as the content and proportion of humic-like and protein-like components of DOM, might exert a crucial influence on the transport of veterinary drugs in the soils.

To evaluate the effects of brackish-water deficit irrigation combined with microbial fertilizer on soil salinization, microbial community structure, and predicted functions, a field experiment was conducted on moderately saline-alkaline soils in the Hetao Irrigation District using Lycium barbarum as the indicator crop. Brackish water irrigation depths were set under microbial fertilizer application (W1: 180mm, W2: 240mm, W3: 300mm, CK: 330mm). It was found that, compared with CK, total soil salinity under W2 was reduced by 56.77% from the fruit expansion to defoliation stages, and Na⁺ was reduced by 55.87% (excluding the flowering stage). During critical yield-formation stages, K⁺, Ca²⁺ and Mg²⁺ were decreased by 45.29%~78.28%. The sodium adsorption ratio (SAR) was decreased by 33.65% and 54.23% at the fruit expansion and leaf drop stages, respectively. In early growth stage, soil pH was increased by 3.63% under W1(P<0.05), whereas no significant differences among treatments were detected thereafter (P>0.05). The bacterial community was dominated by Proteobacteria, Bacteroidetes, Actinobacteria, and Acidobacteria, with Sphingomonas and Pseudomonas as predominant genera. Under W2, the relative abundance of Acidobacteria was increased by 28.09% (P<0.05), and during the fruit expansion stage the relative abundance of Sphingomonas was significantly higher than under CK; this genus was significantly negatively with Cl^- content (P<0.05) and SAR value (P<0.01). Under brackish water irrigation, Bacillus amyloliquefaciens was selectively enriched under W2 (57.6% relative to CK; P<0.05), and accumulation of Paenibacillus mucilaginosus at the peak fruiting stage was promoted; enrichment of both taxa was maintained through defoliation.Functional prediction indicated that chemoheterotrophy and aerobic chemoheterotrophy were the primary community functions, and the abundances of carbon- and nitrogen-cycling functional groups were increased by 8.25% and 36.31%, respectively, under deficit brackish-water irrigation. Redundancy analysis (RDA) indicated that K⁺, Mg²⁺, Cl^-, and Na⁺ were the key drivers of microbial community variation, with a cumulative explanatory rate of 54.65%. The soil quality index (SQI) based on principal component analysis (PCA) indicated stage-dependent irrigation demand: at the flowering stage, W3increased SQI by 14.46%, whereas from the fruit expansion to defoliation, SQI under the W2 increased by 381.11%. On the basis of the SQI-irrigation volume model, an optimal brackish-water irrigation depth of 240~260mm under microbial fertilizer application was proposed.

Pakchoi pot experiment was carried out with polyethylene microplastics (PE-MPs) and lambda-cyhalothrin. The Illumina NovaSeq high-throughput sequencing technology was employed to sequence and analyze the soil bacterial communities on the 1st and 30th days of the experiment. The results demonstrated that PE-MPs significantly inhibited the root length, fresh weight, and plant height of Pakchoi (P<0.05), which were reduced by 39.63%, 32.9%, and 19.54%, respectively, compared to the control. In contrast, lambda-cyhalothrin promoted plant growth, and when applied in combination, it alleviated the inhibitory effects of PE-MPs. On the 1st day of the experiment, both lambda-cyhalothrin and PE-MPs significantly affected the diversity of the bacterial community (P<0.05) and inhibited the growth of some bacterial groups. As the experiment progressed, lambda-cyhalothrin mitigated the inhibitory effect of PE-MPs pollutants on bacterial groups. LEfSe analysis revealed that the indicator bacterial taxa in different treatment groups varied at different times. FAPROTAX functional prediction analysis indicated that, compared to the control, lambda-cyhalothrin may inhibit the abundance of functional groups related to hydrocarbon degradation on both day 1and day 30, while PE-MPs may significantly promote their abundance. However, this promotional effect was attenuated under combined treatment. Lambda-cyhalothrin may significantly increase the abundance of human pathogen-related functional groups, especially at the beginning of the experiment. Interaction network analysis of the top 30 genera in terms of abundance indicated that PE-MPs promoted the symbiotic and cooperative relationships among these genera in terms of abundance at the end of the experiment (day 30), while the combined pollution significantly enhanced such relationships at the beginning of the experiment (day 1). These findings provide a scientific basis for a deeper understanding of the impacts of PE-MPs and lambda-cyhalothrin pollution on the microecological environment of farmland soil.

This study systematically investigated the adsorption mechanisms of heavy metal Tl by layered double hydroxide (LDH). For the first time, CaAl-LDH was applied to immobilize Tl in three typical Tl-contaminated soils-red soil, yellow soil, and black soil. The immobilization effectiveness was assessed based on the ecological and health risks associated with soil Tl, using the Tessier sequential extraction method and an in vitro test method. Results showed that the active sites on CaAl-LDH surface were uniformly distributed, with Tl⁺ adsorption primarily following monolayer adsorption behavior, consistent with the Langmuir model. The key adsorption mechanisms included: (1) synergistic adsorption induced by anion exchange in LDH interlayers; (2) surface adsorption of Tl⁺ on LDH lamella surfaces and edges; and (3) complexation with Tl⁺ following the release of functional groups from the LDH surface. At the relative optimal application rate (5% of soil mass), CaAl-LDH significantly reduced the mobility factor of soil Tl and the relevant ecological risk and crop uptake risk, with an average decrease of 24.1%. Also, it significantly lowered the bioaccessibility of soil Tl and the relevant human health risk, with an average decrease of 30.5% in the gastric phase and 21.3% in the small intestinal phase. These findings provide a scientific foundation and practical reference for developing immobilizing remediation agents for Tl-contaminated soils, thereby contributing to improved soil environmental quality and human health protection.

Based on the InVEST model and integrating the use of the Geodetector and Geographically Weighted Regression (GWR) model, this paper dynamically analyzed the spatio-temporal evolution characteristics and driving mechanisms of habitat quality in Gannan Prefecture at the county scale from 2000 to 2020. The results show that: 1) From 2000 to 2020, habitat quality continuously deteriorated, with an average annual decline rate of 0.0016a^-1. The spatial distribution was relatively stable, showing a pattern of high in the southwest and low in the northwest-southeast. Hotspots were concentrated in Maqu County and Luqu County, while coldspots were distributed in a band or patch form around towns and in areas with dense road networks, mainly consisting of construction land and cultivated land. 2) The mutual conversion of land use types had a negative impact on habitat quality, with a cumulative decline of 0.031 in the habitat quality index. The conversion from forest land and cultivated land to grassland and from cultivated land to forest land were the main types that improved habitat quality. However, the conversion from ecological land to living and production land and from grassland to cultivated land was continuously expanding, with an average annual conversion area of 20.22 and 194.05km² respectively, further intensifying the risk of habitat degradation. 3) Elevation (0.459), average annual temperature (0.433), average annual precipitation (0.349), average annual evaporation (0.345), GDP (0.274), and population (0.269) were the main driving factors of habitat quality differentiation, and the interaction between these factors was nonlinearly enhanced or double-factor enhanced. The interaction between elevation and other factors was the strongest, with a contribution rate of over 45%, which had a decisive influence on the spatial distribution of habitat quality. 4) The impact of different factors on habitat quality in space was significantly different (P < 0.05). Average annual temperature, average annual evaporation, GDP, and population mainly had negative effects on habitat quality, while elevation and precipitation had significant positive effects. Below 3200m, habitat quality was relatively low, ranging from 0.045 to 0.203, with severe degradation fluctuating between 0.146 and 0.232. Above 3200m, precipitation was abundant and habitat quality was relatively high (0.383~0.759), but this area was more sensitive to climate change, with low grassland continuity and point-like habitat degradation. In future environmental protection, strict control over the expansion of construction land should be implemented, the system of land use control for cultivated land should be strictly enforced, and differentiated and diversified regulatory strategies should be adopted considering the characteristics of different driving factors. At the same time, attention should be paid to the impact of climate change on habitat quality in high-altitude areas, and through refined regulation of the human-land relationship and strategies to address climate change, the overall quality and efficiency of habitat should be improved.

This study established an integrated SWAT-LSTM-MLP model to investigate phosphorus flux dynamics in the Baiyangdian Basin, China. The SWAT simulations revealed three key findings: (1) the average annual phosphorus input of the basin was 717.3t/a(57.6kg/km²·a) during 2019~2023, and the organic phosphorus (78.8%) was the main form, which showd an increasing trend in 2023; (2) the phosphorus flux showed significantly seasonal variation, the flood seasons (June-September) accounted for 85.57% of annual phosphorus loading (49.28kg/km²), which was significantly positive correlations (P<0.01) with precipitation, runoff, and sediment load; (3) for the spatial distribution, the phosphorous load of Fu-Cao-Pu River (79.3kg/km², 32.09%) and Baigou Diversion Canal (66.6kg/km², 26.96%) was higher than other region. Furthermore, in order to avoid the deficiency of the SWAT mobile in nonlinear varaitions, the LSTM-MLP hybrid model was applied to improve the accuracy of phosphorous load prediction. The model results showed that:(1) the LSTM-MLP hybrid model significantly improved prediction accuracy (NSE>0.9, RMSE<4kg/km²) compared to conventional LSTM (NSE>0.8, RMSE<7kg/km²), particularly in capturing peak values; (2) simulate analysis under different sediment reduction scenario demonstrated that the Baigou Diversion Canal under 60% sediment reduction scenario can reduce 35.29% of phosphorus load, which was the most effective regions among all subregions.

This paper revealed the spatial and temporal distribution and source characteristics of dissolved organic nitrogen (DON) in the watershed through the joint analysis of water quality indicators, three-dimensional fluorescence spectroscopy (3DEEM) and Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS). The results showed that the proportion of DON concentration to total dissolved nitrogen concentration in all counties of the watershed in all seasons ranged from 15% to 20%. The overall DON level was low, while the DON/TN ratios in the main stream sections of the two downstream counties were significantly higher than those in the upstream. 3DEEM showed that DOM was dominated by humus of terrestrial origin, and the activity of the protein-like fractions was enhanced in summer and autumn. The fluorescence index indicated that agricultural input dominated the exogenous pollution. FT-ICR MS detected that nitrogenous organic matter was dominated by lignin-like substances. The input of Yellow Water (YH) elevated the percentage of condensed aromatics downstream by 11.4%. The lower amount of nitrogenous organic matter in the water body in summer compared with autumn at Gaojiaxie Village (G9) revealed the influence of land-source humus input. The transformation of dissolved organic nitrogen in the Fen River basin was synergistically driven by exogenous inputs, biodegradation and photolysis, which formed the characteristics of DON sources under the action of multiple mechanisms. The study suggested that priority should be given to the control of tributary inputs of difficult-to-degrade organics to enhance the effectiveness of management.

To investigate the pollution sources of heavy metal elements and assess the health risk status in the water of South Dongting Lake, and to provide data and methodological support for the precise prevention and control of heavy metal pollution and public health management, a total of 24 surface water samples and 16 groundwater samples were collected in the area. The absolute principal component score-multiple linear regression (APCS-MLR) receptor model and the health risk assessment (HRA) model were applied to analyze and evaluate the contamination sources of eight heavy metals (As, Co, Cr, Mn, Ni, Pb, Zn, Cd) and their associated health risks in the water bodies of South Dongting Lake. The results revealed considerable variation in heavy metal concentrations. In surface water, only Cr exceeded the standard, with an exceedance rate of 33.33% and an average concentration 4.70 times the standard limit. In groundwater, Cr and Mn exceeded the standards, with exceedance rates of 25.00% and 31.25%, and average concentrations 1.32 and 3.07 times their limits, respectively. The remaining elements complied with the relevant quality standards. Most heavy metals showed significant spatial variation, and Cr contributed the most to the total heavy metal concentration in both water bodies. Based on the single-factor and Nemerow pollution indices, Cr in surface water was at a severe pollution level, while Mn and Cr in groundwater were at severe and moderate levels, respectively; however, the overall pollution level of the regional water bodies was relatively light, with good water quality. Pollution source apportionment identified four primary sources: agricultural activities (29.76%), industrial activities (26.78%), natural geological background (23.57%), and traffic-related sources (19.90%). The health risk assessment indicated that risks from the drinking pathway were greater than those from dermal contact, and adults faced higher risks than children. The combined non-carcinogenic risk for adults was primarily attributed to Cr and Zn via the drinking water pathway. The carcinogenic risk of Cr posed a potential threat to both groups through drinking. Furthermore, Cr originating from industrial and agricultural activities was identified as the major contributor to carcinogenic risk. Therefore, to ensure the environmental safety of the water in South Dongting Lake, it was concluded that pollution sources should be strictly controlled, and the supervision and management of heavy metal pollution risks in the water bodies should be continuously strengthened, with priority given to controlling heavy metal pollution caused by industrial activities and agricultural production.