In predicting the climate change using earth system models, land biogeochemistry is one of the most significant sources of uncertainty, both being itself and through its interactions with other sectors within the earth system. Compared to physical processes (e.g., atmospheric dynamics), land biogeochemistry is relatively weak in its theoretical underpinning, and de facto has not a scaling consistent theory to link the many processes, including soil carbon cycling, plant-soil interaction and plant biogeochemical dynamics. I here contend that the theory of Equilibrium Chemistry Approximation, a first order approximation to the law of mass action from physical chemistry, provides a first-principle based way to model the many different biogeochemical processes both separately and interactively. I will demonstrate this with several examples spanning many orders of spatial scales. At the molecular scale, the ECA theory improves the classical Michaelis-Menten kinetics in modeling single or coupled biogeochemical reactions. At the plot scale, ECA allows a predictive modeling of how soil moisture control soil respiration and organic-mineral interactions. In the plant-soil coupling, ECA enables a mechanistic prediction of plant-soil nutrient competition. And finally, across micro-to-macro scales, ECA is able to upscale fine-scale reactive transport models with only a few parameters. Finally, I will discuss other applications that are enabled by the ECA theory.