A conceptual model was developed for linking abiotic factors to plant community characteristics in abandoned agricultural fields that are becoming forested ecosystems through secondary succession. The model addresses environmental controls on plant communities including soil characteristics, nutrients, moisture availability, and light exposure, and in turn, relates those more direct controls to the underlying geology including both bedrock lithology and geomorphology. Data were collected from six mid-successional fields in south central Indiana and subjected to statistical analyses in an effort to empirically evaluate elements of the conceptual model. Statistical results, primarily based on correlation and regression analyses, indicate that plant density variables were the largest influences on biodiversity. Sunlight and soil moisture related variables were prominent influences on herbaceous plant density, while soil chemistry variables were more significantly correlated with tree and shrub density. Individual species correlated with abiotic variables with varying significance. Soil pH was most influential below a threshold of 5.2, resulting in the presence of acid tolerant species. In other cases, high abundance of resources may have favored specific species. For example, facultative wetland vegetation was present at the field with a perched water table. Heterogeneity in conditions influenced intra-field variation in the vegetation. For example, sunlight had more significant relationships with herbaceous plants at fields with variable canopy cover, and the depth to the water table contributed more strongly to plant community variation where it was most spatially variable. The values of the abiotic variables were linked to geologic setting. Limestone bedrock facilitated heterogeneous water table levels and buffered soil acidity. Soils above siliciclastic bedrock tended to have lower pH, although in some locations a fragipan layer may have mitigated soil acidity through retention of base cations due to slow drainage. The study results have implications for natural resource inventories, environmental impact assessments, and resource management decisions.