Predicting Post mortem Interval using Discriminant Function Analysis of the Succession Pattern of Necrophagous Insects

Abstract

Estimation of postmortem interval (PMI) is an integral part of most death investigations. Succession pattern of insects associated with decomposing animals is one of the methods used for PMl prediction. This study aims to design a model with which PMl may be predicted in Lagos and Kaduna States using the successional pattern of insects on decomposing canions in the two regions. Seventy percent of the data on the succession pattern of insects collected from decomposing pigs were fed into discriminant analysis to generate a classification model while the remaining 30% of the data was used to test the model. Twenty six variables were used to classify time of death into nine groups (0-1,1-2,2-3,3-4,4-5,5-6,6-7,7-8 and over 8 weeks). Discriminant function analysis revealed that empty calliphorid puparium around carrion (0.675), Tenebrionidae (0.598), Histeridae (0.582), adult Calliphoridae (-0.509) and Scarabidae (0.50) were the five most discriminating variables for predicting postmortem interval. The model (LD 1) generated was tested with blind data to time of death with an upper accuracy limit of79. 7%.