Acetoclastic and Hydrogenotropic pathway shift in anaerobic digestion

Fig. 1. Conversion rates and efficiencies across key anaerobic digestion phases: The relationship between reactor configurations (R1: 100% food waste (FW), R2: 75% FW : 25% sewage sludge (SS), R3: 50% FW : 50% SS, R4: 25% FW : 75% SS, and R5: 100% SS) and their corresponding performance metrics. Calculations were based on influent and effluent concentrations, converted to gram COD (Chemical Oxygen Demand), using specific conversion factors for each parameter across reactors.Fig. 2. Temporal dynamics of methanogenic pathway shifts in HOLAnD® reactors: The relationship between pH–pKa differentials and digestion time was assessed using measured pH values and pKa calculated from volatile fatty acid (VFA) concentrations. These trends were analyzed alongside biogas, methane, and carbon dioxide yields from the reactors. Hydrogen concentrations were estimated stoichiometrically based on propionic and butyric acid levels.Fig. 3. Profiles of dissociated and undissociated volatile fatty acids: The speciation of VFAs into dissociated and undissociated forms was computed using daily pKa values and corresponding pH measurements from the reactors. Additionally, the visualization were performed using Python, with custom code applied to the associated datasetFig. 4. Multivariate correlation analysis of physicochemical variables and performance indicators in the HOLAnD® system: A composite matrix illustrating the interrelationships among 16 key process parameters, integrating Pearson correlation coefficients (heatmap) and Mantel test results (network overlay). Pearson correlations were computed using reactor-derived data, while the Mantel test was conducted using the scikit-bio package in Python. Visualization was performed using OriginPro for the Pearson heatmap and Gephi (software) for the Mantel network overlay.Fig. 5. Taxonomic and functional organization of microbial communities: A chord diagram illustrating species-level compositional shifts in the microbial species based on the 16S rRNA metagenomic abundance data. Connecting lines indicate variations in relative abundance, with each microbial taxon distinguished by a unique color. Functional annotation was derived through KEGG-based analysis, while the visualization was generated using Tableau for the chord diagram and Gephi (version 0.10.1) for functional network representation.

Fig. 1. 关键厌氧消化（anaerobic digestion）阶段的转化速率与效能：反应器构型（R1：100%餐厨垃圾（food waste, FW）、R2：75%餐厨垃圾:25%污水污泥（sewage sludge, SS）、R3：50%餐厨垃圾:50%污水污泥、R4：25%餐厨垃圾:75%污水污泥及R5：100%污水污泥）与对应性能指标之间的关联。计算基于进水与出水浓度，转换为克级化学需氧量（Chemical Oxygen Demand, COD），并针对各反应器的对应参数采用专属转化系数完成测算。
Fig. 2. HOLAnD®反应器中产甲烷途径随时间的动态演化：通过实测pH值及由挥发性脂肪酸（volatile fatty acid, VFA）浓度计算得到的pKa值，分析pH-pKa差值与消化时长之间的关联。同时结合反应器产生的生物气、甲烷及二氧化碳产率对该变化趋势展开分析。氢气浓度则基于丙酸与丁酸浓度，通过化学计量法估算得到。
Fig. 3. 解离型与非解离型挥发性脂肪酸分布特征：基于反应器每日测得的pKa值与对应pH测量值，计算挥发性脂肪酸解离为解离型与非解离型的形态占比。此外，可视化工作通过Python完成，针对关联数据集编写自定义代码实现。
Fig. 4. HOLAnD®系统理化变量与性能指标的多变量相关性分析：该分析通过一张复合矩阵展示16项关键工艺参数间的相互关联，整合了皮尔逊相关系数（热图）与Mantel检验结果（网络叠加层）。皮尔逊相关性通过反应器实测数据计算得到，而Mantel检验借助Python的scikit-bio工具包完成。可视化方面，皮尔逊热图采用OriginPro绘制，Mantel网络叠加层则通过Gephi软件实现。
Fig. 5. 微生物群落的分类学与功能组成特征：基于16S rRNA宏基因组丰度数据，绘制用于展示微生物物种水平组成变化的弦图。连接线代表相对丰度的变化，每个微生物类群以独特颜色区分。功能注释通过基于京都基因与基因组百科全书（Kyoto Encyclopedia of Genes and Genomes, KEGG）的分析得到，其中弦图的可视化采用Tableau完成，功能网络展示则使用Gephi（0.10.1版本）实现。