Networked PalYnology Models of Pollen and Human Systems (NYMPHS)
Abstract
Forensic palynology uses pollen to link persons or objects to particular places and times. Several features of pollen make it useful as a biomarkerÑits ubiquity in the environment, its durability, and its predictable distribution in space and time. As a result, pollen can be particularly useful for national security applications. For example, palynology has been used to link movement of bodies in mass graves in Bosnia and could contribute to determining the provenance of hard-to-trace items, such as computers, clothing, and undetonated explosives (where tiny pollen grains will settle into the interior), and paper documents. Pollen grains can also be an important marker of objects moved between locations, such as a laptop transported by terrorism suspects from Yemen, to Denmark, to Michigan. Despite these advantages, forensic palynology, to date, has rarely been applied in practice, largely because it is slow, depends on microscopic identification of pollen by a dwindling number of palynological experts, and often results in identifications of limited taxonomic resolution (e.g., family or genus level). Even if scientists could process a sample and identify that sampleÕs plant species, they lack an effective tool to map pollen sample composition information to a geographic location while accounting for human impacts (e.g., cultivation, irrigation, urbanization, etc.) on pollen distributions. However, DNA barcoding of pollenÑidentifying pollen using DNA markersÑspecifically metabarcoding, which uses next-generation DNA sequencing and extends the technique to mixed-species samples, could transform forensic palynology by improving its speed, taxonomic resolution, and availability. Metabarcoding can identify pollen at high taxonomic resolution from multiple plant species in mixtures, even when pollen is not abundant, with standard read depths using next-generation Illumina sequencing. Further, the development of advanced species distribution models (SDMs) and complimentary mathematical and network models that integrate human effects could resolve the pollen provenance problem by estimating distributions of plant (or animal) species in space and time. Based on this information, forensic scientists could, in theory, geolocate a multispecies pollen sample by constructing a set of advanced SDMs (accounting for human effects) for each plant species present in a sample, then calculating geographic overlaps in that set to create continuous suitability surfaces for geographic provenance likelihood for any sample. The purpose of this research program is to make forensic palynology via metabarcoding a key operational contributor to national security by equipping the DoD with a set of reliable, globally validated, easy-to-use geocomputational tools, mathematical models, and SDMs for geolocating pollen samples. The NYMPHS toolbox will not only transform forensic palynology, but forensic science more generally, allowing scientists to determine the fine-scale geographic provenance of a range of forensic sample types, eventually allowing US military investigators and other related domestic and national security forces to track locations associated with people and objects (laptops, IEDs, illicit trade products, etc.) that threaten US security. Our dynamic geospatial modeling framework could also be expanded beyond predicting pollen provenance to identify potential paths traversed by people and objects through space and time, radically extending the DoDÕs capabilities for understanding global movement, key nodes, links, and optimal sites for military interdiction.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Jun 10, 2019
- Source ID
- W911NF1910231
Entities
People
- Anthony Grubesic
Organizations
- Arizona State University
- Army Contracting Command
- United States Army