The versatility and usefulness of aquatic ecosystem cannot be shoved aside due to environmental relevance, commercial benefits and even educational needs. Over the years, the ecosystem has been over bodied with physical, chemical and biological degradation. These degradations have called into question the quality of aquatic ecosystem water. This research paper has applied matrix laboratory (matlab) to aquatic ecosystem dataset in predicting the quality of these water. It has simulated a matlab aquatic ecosystem model with corresponding interfaces. The simulation catered for training, testing, membership and inference interfaces. The histogram graph portrays the progression of each simulation attributein cognizance with histogram values. With an average training error of 5.63 cutting across thirty (30)epochs and an error tolerance of 0.05 the training was indeed optimal in terms of learning eliminating over fitting and underfitting.