Maggots are incredibly useful to forensic scientists tasked with determining time of death in investigations. Blow flies can detect the presence of a corpse from up to two miles away within only minutes of death. Once they hunt down the location, blow flies lay eggs on and within the corpse minute or hours after their arrival.
But it’s not as easy as just analyzing the collected larvae and maggots. The determination of postmortem interval is complicated by the presence of more than one species of blowflies. Although the insect development lifecycle and timeline for blowfly species is well-mapped, the initial determination of which species is which is critical—and of course, it is nearly impossible to visually distinguish between the species. To compensate, entomologists typically rear the maggots to maturity so that the more distinct morphological features of the adults can be used to accurately identify the species. However, this is time-consuming, resource-intensive, and requires the sample to be alive.
Building on their previous work, researchers from the University of Albany and John Jay College of Criminal Justice have developed an alternative using advanced machine learning and direct analysis in real time-high resolution mass spectrometry (DART-HRMS).
“Our early work demonstrated that accurate species attributions could be made using eggs, larvae, pupae and adult insect samples that were made up of a single species,” lead author Rabi Musah told Forensic. “The present work builds further on these findings, and addresses the issue that entomological samples are sometimes comprised of members of multiple species, which makes it very challenging to assign species identities to the multiple individuals in a sample.”
When entomological samples are acquired, they are usually submerged in aqueous ethanol for storage. The researchers previous study revealed that these aqueous ethanol suspensions exhibit species-specific chemical signatures that can be used to rapidly identify species. Tailoring the approach to mixture suspensions, the researchers created solutions containing two, three, four, five, and six blowfly species.
A hierarchical conformal predictor was applied to a hierarchical classification tree that was trained against the DART-HRMS data. Analysis of the mixtures revealed obvious visual separation of individual species cluster, as well as two-, three-, and four-species clusters. The clusters representing five- and six-species mixtures overlap with one another to some extent, indicating that their mass spectral characteristics are highly similar and may be difficult to distinguish. Analysis of mixture-type clusters at the second level of the hierarchy reveals the cluster constituent species identities.
Maximum accuracy of the mixture type model was 94 percent, and the researchers report confidence limits between 80 and 99 percent for species identification through the conformal predictor.
“As anticipated, the accuracies for the two- to five-member species nodes were lower than for individual species because of the increased chemical complexity of these samples when compared to the single-species samples,” the authors wrote in their paper published in Analytical Chemistry.
Previous proof-of-concept success cases have only achieved single-species identification, and while DNA analysis would be the gold standard for identification of multi-species, Musah says it’s just not practical at this point.
“It requires that the DNA of the various insect species be profiled, so that the profiles of sample unknowns can be screened against known genomes in order to assign species identity. The challenge is that the genomes of the vast majority of forensically relevant insect species have not been mapped, so this approach is not feasible for most investigations. Furthermore, the accomplishment of this task is extremely time consuming,” Musah said.
Ultimately, this study demonstrates a novel and rapid approach to the identification of multiple maggot species in one sample without destroy or disturbing the evidence. Not only can these maggots help establish time of death in investigations, but they can also be retrieved from infested wounds to provide timelines of maltreatment in child, elder and animal abuse cases.
Musah and his team will continue their research in a variety of ways. Analytically, they plan to evaluate the scope of the approach for insect identification in field samples and assess the method’s ability to enable multi-species determination for the egg life stage. Perhaps most pertinent, Musah said the researchers plan to expand the DART-HRMS chemical database to include species that are endemic to specific areas and even create a graphical user interface that crime labs can utilize.
Photo credit: Paul Venter