The airborne fraction of soil (dust) is both ubiquitous in nature and contains localised biological and chemical signatures, making it a potential medium for forensic intelligence. Metabarcoding of dust can yield biological communities unique to the site of interest, similarly, geochemical analyses can uncover elements and minerals within dust that can be matched to a geographic location. Combining these analyses presents multiple lines of evidence as to the origin of dust collected from items of interest. In this work, we investigated whether bacterial and fungal communities in dust change through time and whether they are comparable to soil samples of the same site. We integrated dust metabarcoding into a framework amenable to forensic casework, (i.e., using calibrated log-likelihood ratios) to predict the origin of dust samples using models constructed from both dust samples and soil samples from the same site. Furthermore, we tested whether both metabarcoding and geochemical/mineralogical analyses could be conducted on a single swabbed sample, for situations where sampling is limited. We found both analyses could generate results from a single swabbed sample and found biological and chemical signatures unique to sites. However, we did find significant variation within sites, where this did not always correlate with time but was a random effect of sampling. This variation within sites was not greater than between sites and so did not influence site discrimination. When modelling bacterial and fungal diversity using calibrated log-likelihood ratios, we found samples were correctly predicted using dust 67% and 56% of the time and using soil 56% and 22% of the time for bacteria and fungi communities respectively. Incorrect predictions were related to within site variability, highlighting limitations to assigning dust provenance using metabarcoding of soil.

空气中的土壤（灰尘）部分在自然界中无处不在，并且包含局部生物和化学特征，使其成为法医情报的潜在媒介。尘埃的元条形码可以产生感兴趣地点特有的生物群落，类似地，地球化学分析可以发现尘埃中可以与地理位置相匹配的元素和矿物质。结合这些分析，可以提供有关从感兴趣的物品收集的灰尘来源的多方面证据。在这项工作中，我们研究了灰尘中的细菌和真菌群落是否随时间变化，以及它们是否与同一地点的土壤样本具有可比性。我们将灰尘元条形码集成到适合法医案例工作的框架中（即使用校准的对数似然比），以使用根据同一地点的灰尘样本和土壤样本构建的模型来预测灰尘样本的来源。此外，我们还测试了在采样有限的情况下是否可以对单个拭子样本进行元条形码和地球化学/矿物学分析。我们发现这两种分析都可以从单个拭子样本中生成结果，并发现特定地点特有的生物和化学特征。然而，我们确实发现站点内存在显着差异，这并不总是与时间相关，而是抽样的随机效应。站点内的这种差异并不大于站点之间的差异，因此不会影响站点歧视。当使用校准的对数似然比对细菌和真菌多样性进行建模时，我们发现使用灰尘对细菌和真菌群落的样本预测正确率分别为 67% 和 56%，使用土壤的预测正确率分别为 56% 和 22%。错误的预测与场地内的变异性有关，这凸显了使用土壤元条形码来确定灰尘来源的局限性。