Investigating and Predicting the Amount of Environmental Impact in Breeding Warm Water Fish in Guilan Province using Comparative Neuro-Fuzzy Inductive Inference System

In recent years, life cycle assessment (LCA) approach is turned to be a useful tool for investigating and determining the environmental impacts of agricultural products and food industry, so that in most countries, it is used as a tool for decision-making in agricultural production planning. Considering the fish as an important part of the human protein required, an investigation was carried out on the environmental indicators (impact categories) in the system of warm water production in Guilan province. Data related to the production value of inputs (indirect emissions) and their’s consumption (direct emissions) in ponds were collected using sampled questionnaire and Ecoinvent database. The results of normalization showed that marine aquatic ecotoxicity (MAET), acidification (AC) and Freshwater Aquatic Ecotoxicity (FAET) have the highest amount of environmental pollutants as 5.17×10-7, 1.95×10-7 and 0.98×10-7, respectively. Emissions resulting from the production of electricity (direct emissions) and pollutants released from the use of electricity, chemical fertilizers and manure (indirect emissions) have the highest share of pollution on these indicators. Also, the comparison of the results of ANFIS design methods showed that the fuzzy C-means method with 8 clusters, with higher accuracy and less error, was able to predict the values of environmental impact categories.