Deep neural networks are vulnerable to adversarial examples crafted by applying human-imperceptible perturbations on clean inputs. Although many attack methods can achieve high success rates in the white-box setting, they also exhibit weak transferability in the black-box setting. Recently, various methods have been proposed to improve adversarial transferability, in which the input transformation is one of the most effective methods. In this work, we notice that existing input transformation-based works mainly adopt the transformed data in the same domain for augmentation. Inspired by domain generalization, we aim to further improve the transferability using the data augmented from different domains. Specifically, a style transfer network can alter the distribution of low-level visual features in an image while preserving semantic content for humans. Hence, we propose a novel attack method named Style Transfer Method (STM) that utilizes a proposed arbitrary style transfer network to transform the images into different domains. To avoid inconsistent semantic information of stylized images for the classification network, we fine-tune the style transfer network and mix up the generated images added by random noise with the original images to maintain semantic consistency and boost input diversity. Extensive experimental results on the ImageNet-compatible dataset show that our proposed method can significantly improve the adversarial transferability on either normally trained models or adversarially trained models than state-of-the-art input transformation-based attacks. Code is available at: https://github.com/Zhijin-Ge/STM.

Moxifloxacin (MOX), a widely used novel antibiotic, may pose ecological risks at its actual environmental concentrations, as has been detected in aquatic systems. However, its ecotoxicity to aquatic organisms and regulatory mechanisms of phosphorus in eutrophic aqueous environments are still limited. This study aimed to analyze its physiological and biochemical parameters, including cellular growth, chlorophyll fluorescence, photosynthetic pigments, oxidative stress biomarkers, and metabolomics to elucidate the toxicity induced by environmental concentrations of MOX in Microcystis aeruginosa at different phosphorus levels. The results revealed that the EC₅₀ values of MOX on M. aeruginosa at different phosphorus concentrations were 8.03, 7.84, and 6.91 μg/L, respectively, indicating MOX toxicity was exacerbated with increasing phosphorus levels. High phosphorus intensified the suppression of chlorophyll fluorescence and photosynthetic pigments, while activating the antioxidant enzyme, indicating severe peroxidation damage. Metabolomic analysis showed MOX induced different discriminating metabolites under different phosphorus levels, and perturbed more biological pathways at higher phosphorus concentrations, such as starch and sucrose metabolism, pyrimidine metabolism, and glycerolipid metabolism. This indicates that phosphorus plays an important role in regulating metabolism in M. aeruginosa exposed to MOX. The findings provide valuable information on the mechanisms involved in cyanobacteria responses to antibiotic stress, and offer a theoretical basis for accurately assessing antibiotic toxicity in eutrophic aqueous environments.

Antibiotics can affect the photosynthetic system of Microcystis, potentially altering the balance of carbon and nitrogen, which may influence the synthesis of different microcystin (MC) congeners. However, the regulatory mechanisms by which antibiotics affect the synthesis of various MC congeners in Microcystis remain unknown. In this study, the effects of ciprofloxacin (CIP) on the growth, carbon and nitrogen balance, amino acid composition, mcyB gene expression, and production of different MC congeners were investigated in two toxin-producing strains of Microcystis aeruginosa. The results show that CIP exposure significantly inhibited the growth of both strains, achieving an inhibition rate of 71.75% in FACHB-315 and 41.13% in FACHB-915 at 8 μg/L CIP by the end of the cultivation. The intracellular C:N ratio in FACHB-315 increased by 51.47%, while no significant change was observed in FACHB-915. The levels of leucine, tyrosine, and arginine, as identified and quantified by UPLC-MS/MS, were significantly altered at higher CIP concentrations, leading to a reduction in leucine percentage and a notable increase in tyrosine in both strains, which contributed to a reduction in MC-LR proportion and an increase in MC-RR and MC-YR proportion. Additionally, the expression of the mcyB gene was upregulated by as much as 5.57 times, indicating that antibiotic stress could enhance MC synthesis at the genetic level, contributing to the increased toxicity of cyanobacteria. These findings emphasize the significant role of CIP in the biochemical processes of M. aeruginosa, particularly in MC synthesis and composition, providing valuable insights into the ecological risks posed by antibiotics and harmful cyanobacteria.

Mussels are highly efficient filter feeders, playing a crucial role in managing eutrophication and assessing pollution. Although research on nanoplastic (NP) toxicity in marine organisms is expanding, studies on freshwater species remain limited despite freshwater ecosystems being disproportionately biodiverse and vulnerable to pollutants. Here, we quantified the effects of polystyrene nanoplastics (PS-NPs, 50 nm) at concentrations of 0, 2, 20, and 200 μg/L on different growth stages of the freshwater mussel Cristaria plicata. After a 45-day exposure, PS-NPs at concentrations ≥ 20 μg/L damaged intestinal epithelial cilia in both age groups. Exposure to 200 μg/L PS-NPs significantly increased malondialdehyde levels and decreased superoxide dismutase activity in both groups, with adults showing a significant rise in total protein content and juveniles exhibiting marked increases in respiratory and ammonia excretion rates. Additionally, PS-NP exposure significantly altered the relative abundance of gut microbial phyla, including Proteobacteria, Firmicutes, Verrucomicrobiota, and Bacteroidota, with Fusobacteriota also being affected in adults. Juveniles were more sensitive to physiological changes, whereas adults exhibited greater microbiota shifts in response to PS-NP exposure. Therefore, this study provides new insights into the stage-specific effects of PS-NPs on intestinal integrity and physiological and biochemical health in freshwater mussels, underscoring the need for targeted management strategies to protect freshwater ecosystems.

Ciprofloxacin (CIP), a widely used fluoroquinolone antibiotic, is frequently detected in aquatic environments, raising concerns over its ecological risks. In this study, the submerged macrophyte Hydrilla verticillata was employed to investigate its capacity for CIP removal and the associated ecotoxicological effects. A series of batch experiments were conducted to evaluate plant growth, photosynthetic efficiency, oxidative stress responses, CIP biodegradation pathways, and shifts in epiphytic microbial communities. Results showed that CIP significantly inhibited the growth of H. verticillata, with inhibition rates of plant length and fresh weight reaching 15.8% and 29.7%, respectively, at 5 mg/L. Photosynthetic parameters were severely suppressed. Fv/Fm represented the maximum quantum efficiency of PSII, significantly decreased by 94.3% at 10 mg/L, while chlorophyll a and b contents declined by up to 36.1% and 31.2%, respectively, compared to control. Antioxidant responses showed H. verticillata undergo peroxidation damage. Biodegradation analysis revealed that H. verticillata effectively removed CIP from water, with maximum removal rates of 37% at 1 mg/L and 31% at 5 mg/L under high biomass (4.2 g) conditions. CIP accumulation was higher in stems than in leaves, and three biodegradation intermediates (C306, C263, and C248) were identified, suggesting a degradation pathway involving piperazine ring cleavage, de-ethylation, and deamination. High-throughput sequencing further indicated that CIP exposure reduced bacterial diversity and richness on H. verticillata surfaces, while promoting antibiotic-resistant taxa such as Actinobacteria and Bacteroidota. These findings highlight the potential role of H. verticillata in antibiotic-contaminated water remediation.

A mixed-mode TRL calibration algorithm based on symmetrical reflection standards is proposed for differential devices measurements. According to the passive fixture’s generalized reciprocity and the phase relationship between its common-mode and differential-mode transmission parameters, the sign ambiguity is removed and the fixtures’ generalized transfer matrices are obtained for both homogeneous and inhomogeneous transmission lines. To verify the proposed algorithm, it is applied to de-embed the DUTs’ mixed-mode sparameters from measurement data, and the experimental results are compared with that from full wave simulation. The agreement of the data obtained through both approaches indicates that the proposed technique is valid to calibrate linear differential devices.