Analysis of Arsenic, Cadmium, Chromium, and Lead Levels: Interplay of Climate, Soil, and Human Factors in Urban Gardens

Abstract

Understanding the spatial distribution of heavy metals in urban gardens is essential for mitigating contamination risks and supporting safe urban agriculture. This study examines factors influencing heavy metal concentrations in garden soil across San Antonio, Dallas, and Houston. A total of 215 soil samples, collected at the surface and down to 30 cm, were analyzed using Inductively Coupled Plasma–Mass Spectrometry for nine heavy metals, including arsenic ( As ), chromium ( Cr ), cadmium ( Cd ), and lead ( Pb ), which are elements of environmental and health concern due to their potential toxicity and, in some cases, carcinogenic properties. Over half (55%) of the samples exceeded the Environmental Protection Agency residential soil threshold for As (0.68 ppm). To identify key drivers, we applied both multivariate linear regression and a generalized mixed graphical network model. While regression highlights independent relationships between variables, the network model comprehensively captures nonlinearly interconnected relationships across variables. It revealed strong links between As and total soil carbon, Cd and clay content, Pb and distance to city center, and Cr and proximity to Superfund sites. A random forest model, informed by the top predictors from the network analysis, achieved its best performance for As (). A combined network model including all four metals emphasized distance to the city center and clay percentage as the most influential factors. The highest average concentrations occurred in gardens within five miles of the city center and in soils with <15% clay. These findings highlight the major drivers of metal concentrations in urban garden soils within the study region.