Abstract:Matter and energy exchanges constitute the basis of ecological processes. Therefore, ecosystems can be regarded as open thermodynamic systems, with bio-species as key components. The changes in ecosystem states can therefore be measured and compared in terms of Gibbs free energy (G) and chemical potential (μ). Based on thermodynamic theories, an ecosystem Gibbs free energy model was established for calculating the changes in G of plant communities and μ of the plant species, at an ecological restoration site of manganese-ore tailing in the Xiangtan region. The total area of the experimental site, including the control plot, was 4 hectares, which was covered with Paulownia fortunei and Koelreuteria bipinnata as the dominant, transplanted, wood species. The rooting areas of the wood species at the experimental site were amended with organic manure, containing a tolerant bacterium strain isolated from the experimental site's mining waste. Organic manure application was undertaken not only to provide necessary nutrients for plant growth, but also to improve the root growth conditions by reducing the metal toxicity in the rhizosphere. The wood plant species were screened first, and the amending effects of the organic manure were tested in pot experiments before initiating the on-site ecological restoration project. For comparison purposes, an equivalent amount of chemical fertilizers was applied to the control plot. Within a five-year period after transplanting the wood species to the site, native plant species naturally germinated as well, and grew in both the restoration and control sites. Soil and plant samples were collected and analyzed using standard methods. The number of plant species was counted, and the density of each species was estimated at the site. Furthermore, the cover percentage, biomass quantity, and manganese uptake of each plant species were calculated. The number of plant species at the restoration site was 48, which was 4 times that of the control plot. The total dry biomass and manganese uptake at the restoration site was 23,324 kg/hm2 and 4,280 g/hm2, respectively. This was respectively 20.6, and 2.6 times as high as in the control plot. In addition, the plant community's total Gibbs free energy was significantly higher at the restoration site than at the control site. These results indicate that the application of organic manure improved the root growth conditions of the contaminated soil, and that the effects were significant. There were also significant differences in chemical potentials among the plant species at both the restoration and control sites (P < 0.001). The μ values for different plant species ranged between -3.79 to 6.76, and -3.42 to 3.59, respectively, at the restoration and control sites. The μ value reflects the ability of plant species not only to adapt, but also to restore the metal-contaminated environment. As ecosystem properties, both G and μ are functions of biomass, manganese uptake, density, and the number of plant species. The values of G furthermore, comprise integrated information on the productivity and biodiversity of the ecosystem. The values of μ concern the growth potential, heavy metal accumulation capacity, and relevant ecological characteristics of the plant species regarding their mutual competition, inhibition, and symbiosis behaviors under given site conditions. In comparison with traditionally applied hyperaccumulators and biodiversity indices, the G and μ values were found to be appropriate parameters for evaluating the impacts of a site's environment on ecosystem growth. Furthermore, they proved suitable for analyzing the ecological behaviors of the remediation plants. In sum, both G and μ values can be used as important indices for determining the effects of ecological restoration in heavy metal contaminated areas, as well as for the screening of phytoremediation plant species.