Download RAW file: https://library.csbe-scgab.ca/docs/journal/39/39_2_123_raw.pdf
Artificial neural networks (ANNs) were used to model the acclimation of indigenous microorganisms in soil contaminated with diesel fuel or creosote. Acclimation data were obtained by measuring the appearance of 14C (as C02) mineralized from radiolabelled tracers that were added to soil microcosms. The ANNs were trained and tested with the following inputs: incubation temperature, water content (as percentage of the soil's water holding capacity), addition or not of nitrogen and phosphorus fertilizers, and sampling time. The ANN output was directly related to the cumulative percent of 14C recovered (%L I4C). The resultant ANN models were incorporated into a user-friendly software package called AccliMat, written in QuickBASIC. With this package, the ANN models can be utilized to calculate the amount of%L14C that will be recovered after a given number of days, or the number of days required to reach a given %LI4C.
A.M. Suchorski-Tremblay and R. Kok 1997. ARTIFICIAL NEURAL NETWORK MODELLING OF MICROBIAL ACCLIMATION PERIODS IN SOIL CONTAMINATED WITH PETROLEUM HYDROCARBONS. Canadian Agricultural Engineering 39(2):123-130.
Canadian Society for Bioengineering