Introduction

Medicocriminal forensic entomology focuses primarily on providing evidence of the amount of time that a corpse or carcass has been exposed to colonization by insects, which helps to estimate the post mortem interval (PMI). Specifically, the estimate is of a minimum post mortem interval (PMImin), because death may occur a variable amount of time before colonization (Fig. 7.1); the maximum post mortem interval (PMImax) is estimated using the time that the person was last seen alive. Forensic entomology derives the bulk of its evidence from two sources: the ecological succession of carrion insect communities and the development of immature insects (Byrd and Castner 2001; Catts and Haskel 1990; Smith 1986). This chapter is concerned with assessing the confidence that can be placed in the accuracy of estimates derived from insect development. (Schoenly et al. 1996) dealt with this theme in succession-based estimates of PMI .

A PMImin based on development is estimated by calculating the age of the oldest immature insect on a corpse using various mathematical models (Grassberger and Reiter 2001; Higley and Haskell 2001; Reiter 1984). The most popular of these is the thermal accumulation model (Higley and Haskell 2001), which takes into account linear effects of temperature on species-specific growth rates to enhance its accuracy. Other models of even greater sophistication have been designed for even greater accuracy (Byrd and Allen 2001; Ieno et al. 2010). Models are commonly implemented on computers and their equations can generate a spurious level of precision - eight or more significant figures - that far exceeds the realities of the biology underlying them. For at least this reason, estimates of a PMImn need to be

M.H. Villet (*), C.S. Richards (*) and J.M. Midgley (*)

Southern African Forensic Entomology Research Laboratory, Department of Zoology and Entomology, Rhodes University, Grahamstown, 6140, South Africa e-mail: [email protected]; [email protected]; [email protected]

J. Amendt et al. (eds.), Current Concepts in Forensic Entomology, 109

DOI 10.1007/978-1-4020-9684-6_7, © Springer Science + Business Media B.V. 2010

ID Q

DC 3

Fig. 7.1 Time line summarizing events in a generalized death investigation, and indicating maximum (PMImax) and minimum (PMI ) estimates of the post mortem interval. The grey boxes associated with each interval are their 'windows' of prediction, which may be asymmetrical. The accuracy of the estimates is reflected in how close the windows are to the actual events they estimate, while the precision of the estimates is reflected in the width of the window. The spacing of the events is arbitrary; some events could be practically simultaneous

Fig. 7.1 Time line summarizing events in a generalized death investigation, and indicating maximum (PMImax) and minimum (PMI ) estimates of the post mortem interval. The grey boxes associated with each interval are their 'windows' of prediction, which may be asymmetrical. The accuracy of the estimates is reflected in how close the windows are to the actual events they estimate, while the precision of the estimates is reflected in the width of the window. The spacing of the events is arbitrary; some events could be practically simultaneous framed by a 'window' of prediction (Fig. 7.1) that gives a measure of the precision of the estimate (Catts and Haskel 1990). This window can be estimated from the statistical confidence interval of the model, but it will probably need further qualification based on information about the biology of the relevant insects and the weather conditions around the putative date of oviposition.

Several authors have provided useful introductions to the growing field of factors confounding estimates of PMImin (Campobasso et al. 2001; Catts 1992; Greenberg and Kunich 2002; Higley and Haskell 2001). The following discussion first examines the concepts of precision, bias and accuracy. It then reviews variables that affect the use of insect development to estimate a PMImin, particularly in terms of their likelihood of occurrence and the magnitude of their effects on precision, bias and accuracy, and suggests ways to take them into account. The discussion concludes with some general comments about making estimates based on insect development.

0 0

Post a comment