Multi-Species Inventory of Biomass and Carbon Sequestration Potential of Tree Species across Gujarat’s Regional Ecosystems, India

This dataset presents an exhaustive dendrometric and biochemical analysis of various tree species found in the Western and Northern regions of Gujarat, India. The primary motivation for this collection is the lack of localized allometric data for common regional species such as Madhuca longifolia and Mangifera indica when grown in specific soil-climatic zones like Banaskantha, Junagadh, and Dahod. By providing a direct link between physical parameters (radius and height) and environmental outputs (Total Biomass and CO2 Tonnes), this dataset enables researchers to move beyond generalized tropical forest averages toward high-precision, species-specific modeling. The data was generated through a rigorous field-to-model pipeline designed to minimize measurement error and maximize reproducibility. The dataset consists of 120 records with 15 variables. The data has undergone extensive statistical validation to ensure its utility for predictive modeling and environmental auditing. The dataset reveals a wide spectrum of carbon storage, from saplings sequestering <1 kg to massive mature specimens like Bombex ceiba (Bombacaceae), which recorded a total biomass of over 3.7 million kg. The mean height across the dataset is 20.38 meters, with a mean CO2 sequestration of 215.9 tonnes per specimen. A Pearson correlation analysis confirms a near-perfect relationship between Radius and Carbon Weight (r = 0.94), suggesting that trunk girth is the most reliable field-measurable proxy for carbon content. A Multiple Linear Regression model was developed. The model achieved an R2 value of 0.7737, indicating that 77.4% of the variance in carbon sequestration can be predicted solely from these two physical dimensions. Using the Shapiro-Wilk test, it was determined that the carbon distribution is non-normal (p < 0.05), exhibiting a heavy right tail due to "megatrees”. Consequently, a Kruskal-Wallis H-test was applied to compare sequestration across the top 5 botanical families. The result (p = 0.078) indicates that while family-level differences exist, the variance in age and size within a single family often outweighs inter-family differences in this regional context. This dataset is designed for modular use by various stakeholders, Environmental Researchers, Urban Forestry Planners, Policy Makers and Educational Use. As a "Gold Standard" dataset for teaching biostatistics and allometric modeling in botany and environmental science programs. Every record in this dataset has been audited for mathematical consistency. Data cleaning protocols were used to identify and verify "mega-specimen" outliers, ensuring that the high values recorded (e.g., in the Junagadh and Banaskantha records) are biologically plausible given the reported dimensions. The current dataset is standard for preliminary regional audits, researchers seeking sub-gram precision are encouraged to cross-reference these findings with species-specific specific gravity tables provided in the documentation.