[25]. The EEM plots were generated from the sellckchem fluorescence spectral data using Sigmaplot 10.0 software (Systat Software, Inc.).2.3. Fluorescence Quenching Titration and Complexation ModelingFluorescence quenching titration was carried out to characterize the complexation of manure DOM with Cu according to the research of Plaza et al. [21]. Experiments were carried out by adding 0.01M Cu(NO3)2 solutions to a series of glass bottles that contained 50mL of DOM solution. The pH value was then adjusted to 7.0. All samples were shaken in the dark for 24 hours under a nitrogen atmosphere at constant temperature (25 �� 0.1��C) to ensure complexation equilibrium.The selection of the wavelengths for the fluorescence titration was based on the highest fluorescence intensity observed from the EEM of the samples [21].

The complexation model of Ryan and Weber was used to determine the binding parameters between DOM and Cu ions [26]. The model assumes a simple 1:1 equilibrium between a metal ion and an organic ?(1+KCuCL+KCuCCu)2?4KCu2CCuCL),(1)where??ligandI=I0+(ICuL?I0)(12KCuCL)��(1+KCuCL+KCuCCu I and I0 are the fluorescence intensity (arbitrary units) at the Cu concentration of CCu and at the start of the titration, respectively, ICuL is the limiting value below which the fluorescence cannot decrease with the addition of Cu2+, CL is the total ligand concentration, CCu is the total Cu ion concentration, and KCu is the conditional stability constant.The complexation capacity (CCCu), that is, the amount of active binding sites per unit mass of DOM, was calculated asCCCu=CL(DOM)total,(2)where (DOM)total is the total concentration of DOM.

KCu and CL were solved by a nonlinear regression analysis with the software 1st Opt 1.5 (7D-soft High Technology Inc., China). The optimum set of fitting parameters for each DOM sample was obtained by iteratively varying the adjustable parameter values until the sum of the squares of the differences between the observed and fitted values of I was minimized. Full, unconstrained optimization was achieved using the quasi-Newton algorithm. 2.4. Statistical AnalysisCorrelations were analyzed between percentages of fluorescence response (Pi,n) to elucidate the transformation of DOM during the composting. Meanwhile, correlation analysis was also used to determine the correlations between Pi,n and parameters of DOM binding with Cu (i.e., CCCu and log KCu). Statistical analyses were GSK-3 performed with the software SPSS 11.5 (SPSS Inc., Chicago, IL, USA) for Windows. 3. Results and Discussion 3.1.