The analysis of textiles associated with decomposing remains as a natural training aid for cadaver-detection dogs

Cadaver-detection dogs are employed by law enforcement agencies to locate human remains. The ability of cadaverdetection dogs to locate human remains relies heavily on the use of effective training aids. Cadaver-detection dogs may be trained using a variety of materials ranging from natural scent sources to synthetic materials. Natural scent sources are typically considered to be the most effective training aids; however, there is concern that using individual tissue types as natural training aids may not be indicative of the scent of an intact human cadaver. The objective of this work was to determine how well textiles associated with decomposing remains retain and mimic the odour of natural training aids. To test this, the chemical odour profile of textile samples collected from decomposing porcine remains that were buried clothed in 100% cotton t-shirts was examined. Throughout various stages of decomposition, the pig carcasses were exhumed and cotton samples were obtained. The volatile organic compound (VOC) profile of the textiles was collected using headspace solid phase microextraction (HS-SPME) and analysed using comprehensive two-dimensional gas chromatography – time-of-flight mass spectrometry (GC×GC-TOFMS). This study provides evidence that textiles associated with decomposing remains may represent a useful natural training aid with a VOC profile reflective of a large subset of cadaveric decomposition odour. The odour profile is dynamic and changes over time suggesting that obtaining textiles from different postmortem intervals would be useful for providing training aids that represent the full spectrum of decomposition odour that cadaver-detection dogs may encounter during a search.


Introduction
The innate ability of canines to locate and hunt prey makes them an ideal candidate for use as a scentdetection tool. Both wild and domesticated canines have a natural ability to detect the scent of their prey [1], an ability that can be used to train canines towards almost any desired scent. The use of canines in a forensic investigation dates to the 1800s, when bloodhounds were used in an attempt to locate Jack the Ripper in England [2]. Currently canines are used for the detection of drugs, explosives, agricultural products, accelerants, currency, missing persons, human remains and certain diseases. Cadaver-detection dogs are specially trained canines employed by law enforcement agencies to locate human remains in cases of missing persons believed dead, suspected homicides and following natural or man-made disasters. These canines evolved when handlers observed that the search and rescue dogs, trained to locate living humans, would lose their tracking ability once the individual was no longer alive, causing their scent to change [3]. Cadaver-detection dogs are still currently one of the preferred search methods for the localisation of human remains as they can cover large areas rapidly and can work both day and night [4,5].
The ability of cadaver-detection dogs to detect human remains relies heavily on the use of effective aids during training. Several materials, both natural and man-made can be used when training these dogs on a specific target odour. Natural training aids include biological tissues such as blood, bone or flesh, decomposition fluid or soil that has been in contact with decomposing remains [1]. Although ideal, the use of whole human cadavers for training is uncommon due to the ethical and legal restrictions associated with acquiring bodies. In addition to natural training aids, synthetic scents have also been developed such as Pseudo™ Scents. These man-made scents are easier to obtain as there are fewer ethical restrictions than those associated with natural training aids, and the synthetic scents are easier to store. Despite these advantages, commercially available synthetic scents often have an overly simplistic chemical composition that is not representative of decomposition odour [6,7], and in some cases can also comprise hazardous chemicals. Cadaverine and putrescine, two very odorous compounds often associated with decomposition, are commonly found in synthetic scents [8].
Although these scents are associated with human decomposition, they also result from the decomposition of any organic matter and are thus not human specific [9].
Due to the current inconsistencies associated with synthetic scents, natural training aids are considered more reliable and provide a better representation of the scent of human remains. However, Hoffman et al. [10] demonstrated that the odour produced from individual tissue samples (e.g. blood, muscle, skin, adipocere, fat, bone, teeth, etc.) shared similarities, but varied enough in their odour profile that care should be taken when using a specific individual tissue type as a cadaver-detection dog training aid. The use of individual bones, flesh and blood samples as training aids may not provide an adequate odour representation of a cadaver or intact human remains. An alternative   178  179  180  181  182  183  184  185  186  187  188  189  190  191  192  193  194  195  196  197  198  199  200  201  202  203  204  205  206  207  208  209  210  211  212  213  214  215  216  217  218  219  220  221  222  223  224  225  226  227  228  229  230  231  232  233  234  235  236 Page 5 training aid that may provide a more comprehensive profile than using bone, blood or specific tissue samples is the use of textile samples that have been in contact with decomposing remains. The use of textiles as training aids is beneficial over the use of bones or blood as they can trap scent molecules from the whole body rather than specific tissues types [11]. Textiles are commonly found in association with decomposing remains [10,11], either fully or partially clothed. As the remains decompose the resulting fluid released will cause the clothing to become stained. It has already been established that decomposition fluid will be absorbed into textiles and the fluid composition will change over time [12]. Additionally, fluid generated by decomposing remains can become embedded in-between the fibres of certain textile types, and might thus remain trapped as decomposition odour for longer than the remains. Cadaver-detection dogs are often trained on multiple training aids to ensure that the full range of decomposition odour is accounted for and that the remains can be successfully located.
The decomposition process is initiated the moment the heart stops beating. The process of breaking down the soft tissue to completion may take as little as days or progress slowly for years. Initially the pH of blood will decrease and the skin colour will fade in a process known as pallor [13]. During this process the early insect colonisers such as blow flies (Calliphoridae) and flesh flies (Sacrophagidae) arrive at the corpse [13]. The arrival of insects to the remains demonstrates that a scent is being emitted from the body. This scent will change and amplify as internal bacteria break down the macromolecules in the body. The breakdown causes the release of gases (i.e. volatile organic compounds; VOCs) such as methane and hydrogen sulfide. This large gas accumulation inside the remains will eventually cause the skin to rupture, effectively releasing the gases. During this period the strong odour anecdotally associated with decomposing remains is detected.
The location of human remains greatly impacts the decomposition process. Whether the remains are deposited on the surface or buried changes the immediate surrounding environment and can alter access by insects and other scavengers. Temperature aboveground is generally higher than below the surface [14,15]. Burial environments tend to result in decomposition over a longer period of time due to this temperature difference as well as due to the protection of the remains from scavenging, large and small [13] and general weathering activities [15]. As the process of decomposition in a burial environment is generally slower than a body decomposing on the surface, it is hypothesised that a burial environment might allow the analysis of the VOC profile from textiles for a longer time period.
Headspace solid phase microextraction (HS-SPME) combined with gas chromatography -mass spectrometry (GC-MS) has been used to identify trace amounts of volatile compounds emitted from a variety of forensic specimens [3,16]. A previous study by Zhu et al. [17] resulted in the development of a method used to detect six target VOCs along with the total volatile organic compounds (TVOCs) from textile samples using HS-SPME and GC-MS. The one-dimensional targeted GC-MS analysis was deemed successful; however, would be unlikely to provide sufficient resolution for the analysis of VOCs from textiles associated with decomposed remains. Decomposition odour has previously been reported to consist of a complex mixture of VOCs [18][19][20][21][22][23][24] and it is still not known which VOCs are responsible for the detection and localisation of remains by cadaver-detection dogs. The initial decomposition odour studies also used GC-MS for VOC analysis, however, due to the complexity of the resulting profiles, comprehensive two-dimensional gas chromatography -time-of-flight mass spectrometry (GC×GC-TOFMS) instrumentation was introduced. GC×GC employs a second dimension GC column for further separation of the eluent from the first dimension column, and has been shown to provide increased peak capacity, higher resolution separations and improved sensitivity, providing a more comprehensive VOC profile [25]. The current study used HS-SPME-GC×GC-TOFMS to investigate the VOC profile emitted from textiles buried with decomposing remains over a two year period. The objective was to investigate how well the odour retained in clothing associated with these remains reflected the VOC profile of decomposition typically reported in the literature for humans and human analogues. This information will provide further insight into the value of textiles as a training aid for cadaver-detection dogs.

Experimental Design
Field experiments consisted of burying and subsequently exhuming a total of seven pig carcasses (Sus scrofa domesticus L.) at various intervals over a 24 month period (January 2013 -January 2015). Pigs were chosen for this study as they are widely accepted as human decomposition analogues in taphonomic studies due to their similarity in internal anatomy, fat distribution, size of chest cavity, skin, gut flora and lack of heavy fur [26]. Pigs were purchased postmortem as excess stock from  Cotton was chosen for this study as it represents a textile that is commonly worn and often associated with human remains. Seven cotton t-shirts were also buried in individual graves in the absence of remains to serve as control samples. All graves were dug a minimum of 3 m apart (with a minimum of 5 m between the control and experimental graves) to a depth of 50 cm using an excavator and were backfilled using an excavator. The clothed pig carcasses and control textiles were exhumed after 1, 3, 6, 12, 18 and 24 months post-burial.
After exhumation, the textile samples were packaged into small paper envelopes, placed into individually labelled paper bags and stored in a cooler for transportation to the laboratory. Paper was selected as the packaging material in order to prevent the textile samples from storing moisture and degrading during transport and storage. To prevent bacterial and fungal growth the textile samples were air-dried under ambient temperature by hanging vertically inside a laboratory fume cupboard, and any adhering tissue, soil or hair was removed after drying. Using sterilised scissors, triplicate 1 × 1 cm cotton samples were collected from each experimental and control textile. Triplicate 1 × 1 cm cotton samples were also collected from a t-shirt prior to burial (referred to as day 0 control textiles) for the purpose of determining background VOCs associated with the cotton textiles. The scissors were washed with acetone between each replicate and between experimental and control samples. The 1 × 1 cm cotton samples were placed into individual 20 mL headspace vials, sealed airtight with a screw cap containing a 1.3 mm thick polytetrafluoroethylene/ silicone septum (Sigma-Aldrich, Castle Hill, NSW, Australia) and stored at -18 °C prior to analysis. VOC profiles of the cotton samples were collected using HS-SPME and analysed using GC×GC-TOFMS.

HS-SPME Sample Collection
Based on previous literature examples that used SPME for the headspace collection of VOCs produced from textiles [27] and decomposing remains [3,10,[28][29][30], two different SPME fibres were chosen for testing using the experimental textile samples collected after 6 months post-burial:  355  356  357  358  359  360  361  362  363  364  365  366  367  368  369  370  371  372  373  374  375  376  377  378  379  380  381  382  383  384  385  386  387  388  389  390  391  392  393  394  395  396  397  398  399  400  401  402  403  404  405  406  407  408  409  410  411  412  413   Page 8 blank was completed before sampling and after every 3 sample collections. Prior to sample collection, the SPME fibre was pre-loaded with an internal standard [31] by exposing the fibre to the headspace of a 200 µL solution of 100 ppm bromobenzene (GC grade, Sigma-Aldrich), prepared in methanol (HPLC grade, Sigma-Aldrich), inside a sealed headspace vial for 15 s at room temperature. After incubating the sample vial for 10 min in a dry bath heating block (Thermoline Scientific, Wetherill Park, NSW, Australia) at a constant temperature of 40 °C, sample extraction was performed by exposing the SPME fibre to the headspace within the sample vial for 10 min while the sample vial was continuously maintained at 40 °C. Australia) was used to connect the 1 D and 2 D columns. The Rxi ® -624Sil MS column is recognised as being highly selective for VOCs. The suitability of the Rxi ® -624Sil MS × Stabilwax ® column combination has been assessed in a previous study for decomposition VOC profiling [32].

GC×GC-TOFMS Sample Analysis
Sample introduction was performed by desorbing the SPME fibre directly in the GC×GC inlet at a temperature of 250 ºC for 5 min using splitless injection with a 30 s inlet purge time. High purity helium (BOC, Sydney, NSW, Australia) was used as the carrier gas at a constant flow rate of 1.0 mL/min. The 1 D oven temperature program included a 5 min hold at 35 °C followed by a temperature ramp to 240 ºC at 5 ºC/min before a final hold at 240 ºC for 5 min (i.e. a total run time of 51 min). Relative to the 1 D oven, the 2 D oven and modulator were programmed to have a constant offset of +15 ºC and +20 ºC, respectively. A modulation period of 5 s was used with a 1.00 s hot pulse time and a 1.50 s cool time between stages. The transfer line connecting the GC×GC with the TOFMS was maintained at 250 °C. The TOFMS detector was operated at a rate of 100 Hz between m/z 29 -450. The ion source temperature was 200 °C and the electron ionisation energy was 70 eV. The detector voltage was set at +200 V above the optimised detector voltage, which was determined daily prior to sample analysis.

Data Processing
Data processing was performed using ChromaTOF ® (version 4.51.6.0; LECO). The baseline was automatically smoothed with an 80% offset. The expected peak widths in the 1 D and 2 D were set at 30 s and 0.15 s, respectively, based on the typical widths observed for the narrowest, non-saturated peaks within the resultant chromatograms. A minimum signal-to-noise ratio (S/N) of 250 was used for 414  415  416  417  418  419  420  421  422  423  424  425  426  427  428  429  430  431  432  433  434  435  436  437  438  439  440  441  442  443  444  445  446  447  448  449  450  451  452  453  454  455  456  457  458  459  460  461  462  463  464  465  466  467  468  469  470  471  472   Page 9 base peak detection with a minimum of 2 apexing masses and sub-peak detection was cut-off at a minimum S/N of 20. A 65% mass spectral match was required to combine subpeaks with each other and their corresponding base peak. Peak identification was performed by a forward search to the 2011 National Institute of Standards and Technology (NIST) mass spectral library database with a minimum similarity match of >80%.
Chromatographic alignment, normalisation and Fisher ratio computation were performed using the Statistical Compare software feature within ChromaTOF ® (version 4.51.6.0; LECO). Samples were input into Statistical Compare and separated into two classes: control textiles (n = 21) and experimental textiles (n = 18). During chromatographic alignment, peak researching was performed with a minimum S/N cut-off of 20 in order to search for peaks not found during the initial peak finding step. A maximum retention time difference of 10 s (i.e. 2 modulation periods) in the 1 D and 0.6 s in the 2 D was permitted during alignment to allow for retention time deviations between samples. In order for peaks to be identified as the same compound across chromatograms during alignment a mass spectral match >600 was required. Analytes that did not meet this mass spectral match threshold were removed from the final compound list. In addition, analytes were only retained in the final compound list if detected in at least two or more of the samples within a class. Following alignment, the analyte peak areas (calculated using unique mass) were normalised against the bromobenzene internal standard peak area. Fisher ratio filtering was performed in order to identify class-distinguishing compounds based on its success in previous studies [23,[33][34][35][36][37][38]. Analytes with a Fisher ratio (i.e. the ratio of between-class variance to within-class variance) above the critical value (F crit = 4.11) were exported as a *.csv file and imported into Microsoft Excel for further analysis. The critical value was computed in Microsoft Excel using the F-distribution based on the number of classes in the analysis, the degrees of freedom for each class and the significance level chosen (α = 0.05). Chromatographic artefacts (e.g. column and fibre bleed) were manually removed in Microsoft Excel and the remaining compounds were sorted into one or more of the following chemical classes: alcohol, aldehyde, aromatic, carboxylic acid, ester, ether, halogenated, hydrocarbon, ketone, nitrogen-containing, sulfur-containing or "other" (i.e. compounds with functional groups that did not fit into any of the previously described chemical classes).

Buried Remains
The pig carcasses buried clothed in 100% cotton t-shirts were exhumed after 1, 3, 6, 12, 18 and 24 months of burial for the observation and collection of textile samples. After one month, when the first pig grave was exhumed, a large amount of soft tissue was present and the entire carcass could be easily extracted as a single, intact specimen. The surrounding soil was very moist, even up to 40 cm above the remains. Three months post-burial the remains of the second pig carcass were exhumed intact with a large amount of soft tissue remaining. Again the surrounding soil was still wet. After lifting the remains, the soil beneath was very dark in colour, appearing almost black. After six months post-burial, the limbs of the pig carcass were fully skeletonized while the soft tissue remaining on the torso had begun to break down and a large presence of white adipocere was observed. The grave exhumed after 12 months contained significant amounts of adipocere, with large tissue sections that were well preserved. The remains after 18 months post-burial were very dry and demonstrated a large amount of tissue loss, with mostly skin and bone remaining. After 24 months, the remains were fully skeletonized and the grave environment was very dry.

Textile Damage
The experimental textile sample exhumed with the first pig carcass after one month burial was very discoloured with orange and brown staining. Large sections of tissue adhered to the surface of the t-shirt and a strong odour was produced. The experimental textile from the grave after three months burial attracted a large amount of flies after it was removed from the grave. After 6 ( Fig. 1a) and 12 months of burial, the experimental textiles located underneath the pig carcasses were very well preserved. The experimental textile associated with the pig carcass exhumed after 18 months burial had only a small portion of the textile recovered (Fig. 1b). This remaining section was found to be well preserved and covered in several layers of tissue, as was observed during the previous exhumations after 6 and 12 months of burial. On the final sampling day, 24 months post-burial, the experimental textile was almost completely disintegrated; however the seams and small sections of textile attached to the seams still remained (Fig. 1c).  532  533  534  535  536  537  538  539  540  541  542  543  544  545  546  547  548  549  550  551  552  553  554  555  556  557  558  559  560  561  562  563  564  565  566  567  568  569  570  571  572  573  574  575  576  577  578  579  580  581  582  583  584  585  586  587  588  589  590 Page 11 The severe degradation of textile observed in the last two experimental graves (i.e. 18 and 24 months post-burial) was most likely due to the distinct difference in moisture detected during these two postburial intervals when compared to the other graves. The increased moisture content preserved the remains, especially in the 12 month graves; this was evident by the production of adipocere. It is hypothesised that the presence of remains with soft tissue resulted in the preservation of the experimental cotton samples in these graves.
Control textile samples consisting of 100% cotton t-shirts were exhumed after 1, 3, 6, 12, 18 and 24 months post-burial. The control textile samples were found to degrade rapidly in the soil grave in the absence of any remains and after 12 months there was virtually no part of the t-shirt left other than the seams. The experimental textile samples on the other hand, showed a great deal of preservation suggesting that the presence of decomposing remains prevents the natural bacterial consumption of the cotton fabric. These very distinct visual differences demonstrate another benefit of the use of textiles as training aids. The findings show that textile samples are still likely to be present (and therefore recoverable by scene of crime officers) several years postmortem, even when the remains have skeletonized.

VOC Profile
The VOC profile of the textile samples was collected using HS-SPME and analysed using GC×GC-TOFMS.  591  592  593  594  595  596  597  598  599  600  601  602  603  604  605  606  607  608  609  610  611  612  613  614  615  616  617  618  619  620  621  622  623  624  625  626  627  628  629  630  631  632  633  634  635  636  637  638  639  640  641  642  643  644  645  646  647  648  649 Page 12 PCA was used to reduce data dimensionality and to provide a visual representation of the multivariate structure of the data using scores and loadings plots. A total of 297 analytes were chosen for submission to PCA. Analytes were chosen using the ChromaTOF ® Statistical Compare software feature and Fisher ratio filtering (described in Section 2.4) in order to identify analytes present at significantly different concentrations/levels between the experimental and control textiles. In this case ( Fig. 3), 43% of the variation within the dataset was captured within the first two principal components. Inspection of additional principal components (i.e. third and fourth principal components) revealed very little variation in the dataset (i.e. 7% and 6%, respectively), and were concluded to provide no further discriminatory information.
The scores plot (Fig. 3a) revealed discrimination between the experimental and control textiles horizontally along the first principal component (PC-1) for the majority of the study period, accounting for 31% of the variation within the dataset. Variation was also observed along PC-1 between the experimental textile samples collected at different postmortem intervals and was a result of numerous compounds detected in the profile rather than a few individual VOCs (Fig. 3b). While the experimental textile samples were separated horizontally along PC-1, the control textile samples were spread out displaying variation vertically along the second principal component (PC-2) axis, accounting for 12% of the variation within the dataset. This variability in the control textiles is a result of the interaction between the cotton textiles and microorganisms in the burial environment (e.g. fungi and bacteria in the soil) which produce enzymes that are capable of attacking the molecular structure 650  651  652  653  654  655  656  657  658  659  660  661  662  663  664  665  666  667  668  669  670  671  672  673  674  675  676  677  678  679  680  681  682  683  684  685  686  687  688  689  690  691  692  693  694  695  696  697  698  699  700  701  702  703  704  705  706  707  708 of the textile fibre [41,42]. This destruction of the textiles can lead to changes in colour or staining of the textile and unpleasant odours [41,43].  Table A-1) that were identified to discriminate the experimental textiles from the control textiles (i.e. decomposition-related VOCs).
The VOC profile produced from the experimental textiles collected 24 months postmortem did not appear to be differentiated statistically (Fig. 3a) or chromatographically (Fig. 4) from the control textiles. The authors recognize the low number of replicates in this study (i.e. n = 1 grave per postmortem interval) is a limitation and that the result obtained at 24 months could be an anomaly arising due to differences in decomposition/preservation of both the remains and the textile. As noted   709  710  711  712  713  714  715  716  717  718  719  720  721  722  723  724  725  726  727  728  729  730  731  732  733  734  735  736  737  738  739  740  741  742  743  744  745  746  747  748  749  750  751  752  753  754  755  756  757  758  759  760  761  762  763  764  765  766  767   Page 14 previously, the remains exhumed 24 months post-burial were fully skeletonized and the grave environment was very dry with severe degradation to the textile observed. At this stage of decomposition (i.e. skeletonization) the remains themselves would not be expected to produce an odour as strong as remains that are still actively decomposing (i.e. bloat, active decay and advanced decay stages). Although caution is taken when interpreting these results due to the lack of replicates, it is possible that decomposition odour may not remain trapped within the 100% cotton textiles indefinitely, and that as the postmortem interval increases (and decomposition progresses towards skeletonization) the decomposition odour profile within the textile diminishes. Regardless, these results demonstrate the importance of performing chemical analysis to verify that textiles recovered from remains have retained an odour reflective of cadaveric decomposition before implementing the textile as a training aid in cadaver-detection dog training. Insight into the chemical differences between the control and experimental textiles was gained by examining the loadings plot displayed in Fig. 3b. Those analytes that appear grouped in the left-hand quadrants of the loadings plot (Fig. 3b) were considered to strongly influence or contribute to the placement of the experimental textile samples in the left-hand quadrants of the scores plot (Fig. 3a).
Together, these 231 analytes (circled in the loadings plot (Fig. 3b) and listed in Table A-1) were identified as the VOCs contributing to the discrimination of the experimental textiles from the control textiles and were therefore considered as decomposition-related VOCs. More than half (i.e. ~55%) of 768  769  770  771  772  773  774  775  776  777  778  779  780  781  782  783  784  785  786  787  788  789  790  791  792  793  794  795  796  797  798  799  800  801  802  803  804  805  806  807  808  809  810  811  812  813  814  815  816  817  818  819  820  821  822  823  824  825  826   Page 15 these VOCs have been previously identified in at least one other decomposition odour study published in the literature for human remains or human analogues (see Table A-1). The remainder of the discussion in this article will focus on these 231 analytes that make up the decomposition odour profile detected in the experimental textiles investigated herein. is also displayed with the carboxylic acid class removed (Fig. 5b).
Sulfur-containing compounds (specifically polysulfides) are currently the most widely recognised group of compounds within the decomposition odour profile, as they are the most consistently reported across studies [18,19,[44][45][46][47][48][49]. These sulfur-containing compounds are commonly reported in the earlier stages, especially in the bloat stage [19,45,46]. However, research conducted in the same location as the present study found that polysulfides were detected throughout all decomposition stages in the air above pig carcasses that were allowed to decompose naturally on the soil surface [25]. In the current study, sulfur-containing compounds were not detected as a major contributor to the decomposition odour profile (Fig. 5b), rather the sulfur-containing compound class was the least abundant compound class detected at all post-mortem interval investigated (with the exception of ethers, a rarely reported class of decomposition VOCs). Sulfur-containing compounds were detected mostly in the three month samples, at this stage there was a large amount of tissue present, however, the skin had ruptured and the soil beneath the pig was stained black, thus the likelihood of seeing the sulfur-containing compounds were higher than in the one month samples where the pig carcass was fully intact with less purging of fluids into the soil. Beyond three months the remains were in a later 827  828  829  830  831  832  833  834  835  836  837  838  839  840  841  842  843  844  845  846  847  848  849  850  851  852  853  854  855  856  857  858  859  860  861  862  863  864  865  866  867  868  869  870  871  872  873  874  875  876  877  878  879  880  881  882  883  884  885   Page 16 stage of decomposition and the sulfur-containing compounds may have dissipated and diffused through the soil. Although this compound class may be present during all stages of decomposition, sulphur-containing compounds were not detected in great abundance in the textile samples. These compounds might therefore be less retained by the textiles during burial decomposition and might be preferentially retained in the soil instead. Forbes and Perrault [21] found that VOC samples taken from the soil beneath decomposing remains contained a larger number of sulfur-containing compounds than air VOC samples taken above the same remains, again demonstrating the soil's ability to capture and trap these compounds.  886  887  888  889  890  891  892  893  894  895  896  897  898  899  900  901  902  903  904  905  906  907  908  909  910  911  912  913  914  915  916  917  918  919  920  921  922  923  924  925  926  927  928  929  930  931  932  933  934  935  936  937  938  939  940  941  942  943  944   Page 17 VOCs characterised as esters were initially high before decreasing after the first 6 months. Similar results were reported by Forbes and Perrault [21], who found that when the VOC profile from soil samples was used to distinguish the stages of decomposition, the PCA plot demonstrated that esters were common in the earlier stages of decomposition, and most abundant during active decay. Esters were found to be prevalent in the textile samples when the remains still had a presence of tissue. A large increase in esters was identified in the 12 month samples, which could be due to the presence of adipocere in the grave environment [10]. However, due to the lack of replicate pigs the spike in esters could also be an anomaly and may simply be due to the initial fatty acid content in the pig from that specific gravesite.
Nitrogen-containing compounds were found to increase steadily the first 6 months, before decreasing.
This corresponds to other studies where nitrogen-containing VOCs are known to appear in the early stages of decomposition [47].
Aromatic VOCs (especially phenols) have commonly been reported in decomposition VOC studies [18,19,29,46]. The abundance of aromatics was high for the samples collected within the first 12 months and low in the 18 and 24 month samples. These findings correspond to previous research as aromatics are found in the active decay stage and early stages of advanced decay [45,46], the absence of these compounds are then indicative of samples that are from a late stage of decomposition. When looking solely at phenol, it was found in all samples except the 24 month samples (Table A-1), which again corresponds to previous findings.
Aldehydes are found more commonly in the later stages of decomposition [18,19,46] and peak during the transition from advanced decay to dry remains for surface studies. However, as the current study involved buried specimens, which tends to promote the formation of adipocere [50], it is more likely that the presence of aldehydes would be less significant as these are formed in aerobic conditions. The one month post-burial grave was disturbed by scavengers on day 2 after burial, which resulted in the remains being exposed. However, the remains were covered again and wire mesh was added to avoid further burrowing. This soil disturbance likely introduced air in the grave. Maggots were also found upon excavation and this might explain the presence of aldehydes in the one month textile samples as some aerobic decomposition was likely to have occurred. The remaining sampling months had a relatively consistent and low abundance of aldehydes.
Ethers are a rarely reported class of decomposition VOCs [51], which is consistent with the low number and abundance of ethers detected at all postmortem intervals investigated throughout this study. Hydrocarbons were most abundant in the samples collected 1 month post-burial, before decreasing and remaining fairly stable. Ketones were found to increase for the first 6 months before a decrease was seen in the 12 month samples; the abundance of ketones then increased and remained Page 18 consistent for the last two excavation times. There was a spike in halogenated compounds detected in the 6 month post-burial samples, and a higher amount in the 18 and 24 month samples compared to the other sampling days. Lastly, alcohols were determined to be higher in abundance in samples where soft tissue was still present, similarly to aromatic compounds, carboxylic acids, esters and aromatics.
Overall, there was a lack of sulfur-containing compounds detected in the textile samples; however, this may be due to the fact that the remains were buried in a soil environment rather than being placed on the soil surface during decomposition. Despite the reduced amount of sulphur-containing compounds detected, the remaining compound classes commonly reported in decomposition odour research (i.e. carboxylic acids, esters, nitrogen-containing compounds and aromatics) all produced VOC profiles from the textile samples that were consistent with that of decomposing remains. This demonstrates the ability of textiles to retain an odour comparable to that of decomposing remains and suggests that although the profile may not be retained indefinitely, as determined by the PCA analysis ( Fig. 3a), textiles could be a viable option as a training aid for cadaver-detection dogs.

Conclusions
This study provides evidence that 100% cotton textiles associated with decomposing remains may represent a useful natural training aid for cadaver-detection dogs with a VOC profile reflective of a large subset of cadaveric decomposition odour. Obtaining textiles associated with decomposing remains from different postmortem intervals is useful for providing training aids that represent the broad spectrum of decomposition odour cadaver-detection dogs are likely to encounter in the field when searching for remains at varying stages of decomposition. Results suggest that decomposition odour may not remain trapped in the 100% cotton textile indefinitely, and therefore chemical analysis is valuable for the verification that textiles recovered from remains have retained an odour reflective of cadaveric decomposition before the textiles are employed as a training aid. Chemical analysis could likewise prove useful in confirming that the odour is retained overtime with prolonged use and storage.