A Method for Reconstruction of Unmeasured Data on Seasonal Changes of Microorganisms Quantity in Heavy Metal Polluted Soil

Abstract

The article presents results of application of the hybrid combinatorial-genetic algorithm COMBI-GA to building models simulating the dependence of quantity of microorganisms in soil on the meteorological conditions and concentration of a heavy metal in an experimental plot. The models built on the rarely measured data during the vegetation seasons are used then for reconstructing the unmeasured decade data on seasonal changes of microorganisms quantity in the soil of a polluted plot during the whole season taking into account the complete support series of the decade meteorological data. This method is demonstrated on the results of modelling amylolytic microorganisms quantity dependence on measured weather factors and concentration of copper in the soil of experimental plots. Meteorological data included the humidity and temperature of air of the current and previous decades. Linear and nonlinear models of changing the microorganisms quantity in control and polluted plots are build based on the rarely measured data during the vegetation seasons. Nonlinear models are used for reconstructing the unmeasured decade data taking into account the complete support series of the decade weather data. Such a methodology can reduce in the future the cost of expensive and time-consuming experiments. A generalized model of amylolytic microorganisms quantity dependence on copper concentration and weather factors is created for predicting critical ecological situations.