Curvature with temperature is a defining characteristic of enzyme catalysed rates. Historically, curvature has been described by Arrhenius behaviour up to a temperature optimum (Tₒₚₜ), while decreases in rates above Tₒₚₜ have been attributed to protein denaturation. However, various evidence indicates that denaturation is insufficient to explain the decreases in rates above Tₒₚₜ. Macromolecular rate theory (MMRT) has been proposed as a description of the temperature dependent curvature of enzyme rates based on the expansion of established thermodynamic principals. MMRT postulates that it is the unusually large change in heat capacity associated with enzyme catalysis (ΔCǂₚ) that results in curvature of rates with temperature. This description is commensurate with the classical description of enzyme catalysis which involves the tighter binding of the transition state (TS) to the enzyme. Here, the molecular origins of the ΔCǂₚ associated with catalysis are explored through crystallography and molecular dynamics simulations. Methods are developed which calculate a ΔCǂₚ commensurate with experimental values from in silico data. Simulations reveal that global rigidification is responsible for reduced heat capacity at the TS, although some protein regions contribute more significantly to differences at the TS. Exploration of the relationship between enzyme mass and catalytic efficiency implicates the vibrational modes captured in dynamics simulations as a reservoir for energy, some of which is available to drive catalysis. This provides a theoretical framework for the large size of enzymes and extraordinary rate enhancements enzyme catalysts achieve. Curvature with temperature is also a feature of organism growth rates and the fluxes through ecosystems. Here, an experimental validation of the applicability of MMRT to these multi-enzyme systems is presented based on the characterisation of an enzymatic pathway. The inherent temperature dependence of an enzyme pathway is found to be an average of the temperature dependence of the constituent enzymes. This relationship shows the factors influencing the rates of enzymes are relevant at the organism level, providing an intrinsic temperature response upon which regulatory responses act. This proof gives an experimental justification of the application of MMRT to organism and ecosystem temperature data.