Soil Fungal Communities Investigated by Metabarcoding within Simulated Forensic Burial Contexts

doi:10.21203/rs.2.22871/v1

Download PDF

Research

Soil Fungal Communities Investigated by Metabarcoding within Simulated Forensic Burial Contexts

https://doi.org/10.21203/rs.2.22871/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 24 Jul, 2020

Read the published version in Frontiers in Microbiology →

Version 1

posted

You are reading this latest preprint version

Background

One of the most debated questions in forensic science is the estimation of the post-mortem interval (PMI). Despite the large amount of research currently performed to improve the PMI estimation, there is still the need for additional improvements, particularly in cases of severely decomposed buried remains. A novel alternative to the morphological examination of the remains is the analysis of the soil microbial communities. Bacteria and fungi are ubiquitous and can be found in the soil and in/on the corpses, and their shifts in populational compositions present at different PMIs may reveal insights for PMI estimation. Despite it already having been revealed that bacteria might be good candidates for this type of analysis, there are knowledge gaps for this type of application when dealing with fungal communities. For this reason, we performed the metabarcoding analysis of the mycobiome present in the soil after prolonged decomposition times, from one- to six-months, targeting both the Internal Transcribed Spacer (ITS) 1 and 2, to elucidate which of the two was more suitable for this purpose.

Results

Our results showed a decrease in the fungal taxonomic richness associated with increasing PMIs and the presence of specific trends associated with specific PMIs, such as the increase of the Mortierellomycota taxa after four- and six-months post-mortem and of Ascomycota particularly after two months, and the decrease of Basidiomycota from the first to the last time point. We have found a limited amount of taxa specifically associated with the presence of the mammalian carcasses and not present in the control soil, showing that the overall the taxa which are contributing the most to the changes in the community originate from the soil and are not introduced by the carrion, extending the potential to perform comparisons with other experimental studies with different carrion species.

Conclusions

This study has been the first one conducted on gravesoil, and sets the baseline for additional studies, showing the potential to use fungal biomarkers in combination with bacterial ones to improve the accuracy of the PMI predictive model based on the shifts in the soil microbial communities.

General Microbiology

Post-mortem interval

decomposition

fungi

metabarcoding

ITS1 and ITS2

taphonomy

Metabarcoding analysis of soil microbial communities has been the subject of numerous studies in forensic science, particularly due to the advent of next-generation sequencing technologies that allow the characterisation of the soil microbiome for numerous applications in a relatively cheap and quick manner[1]. Metabarcoding refers to the characterization of operational taxonomic units (OTUs) by targeting specific “barcodes” in the microbial genome[2, 3], either the bacterial one (selectively amplifying the 16S rRNA genes)[4] or the fungal one (targeting the ITS genomic regions)[5]. This approach usually involves DNA extraction from soil, amplification via polymerase chain reaction (PCR) of a genomic region of interest (e.g. the “barcode” region), sequencing of the resulting amplicons, bioinformatic processing of the information and taxonomic evaluation of the analysed samples[6].

Metabarcoding analysis of microbial communities has recently started to attract the attention of the forensic community as a new potential means of estimating the time elapsed since death, also known as post-mortem interval (PMI)[7–9]. PMI estimation is a key factor in forensic investigations as it can provide important information on the timing and hence circumstances of death; however numerous taphonomic factors[10] play a significant role in the decomposition of organic remains, making the correct estimation of the PMI more challenging and limited in its accuracy, particularly in the case of heavily decomposed remains[11]. For this reason, there is space for the development of additional tools to improve the estimation of the PMI, that could complement existing ones (for an overview of the methods available for PMI estimation see Metcalf[7]). Recently, several independent studies have demonstrated the applicability of this type of study to the estimation of PMI and have investigated the effects of different hosts (e.g. decomposing human[12–16], swine[17–19] or rodents[14, 18, 20]), different burial conditions (e.g. exposed on the surface[12–18, 20] or shallow burials[15, 19]), different environments (e.g. indoor laboratory environments[14, 20] or outdoor environments in different geographical locations[12–19]) and different sources (e.g. in/on the carcass[12, 14, 17, 18, 20] or in the cadaver-associated soil[12–16, 19, 20]). In general, it has been shown that, despite the obvious differences between different experimental designs, some microbial taxa are regularly associated with specific decomposition stages, and that specific taxa appear or increase specifically at selected time points post-mortem in a similar manner across several independent studies[7, 19]. This motivates additional research to further explore the possibility of applying this methodology in future to the justice system, as well as/such as applying machine-learning techniques to assess the error rates of the proposed methodology[7].

According to Metcalf[7], the use of microbes to estimate PMI is particularly promising and may have particular utility when the body is heavily decomposed (e.g. dry or skeletonised remains) due to the current difficulty in the estimation of a precise PMI for skeletonised remains. However only in a few cases were long time frames investigated via microbial metabarcoding[14–16, 19], leaving gaps in current knowledge.

Most previous studies have focused principally on the bacterial communities, despite some of them showing the potential of investigating both the bacterial and the fungal communities[13, 14] to maximise the success of addressing PMI from a microbial perspective. Parkinson et al.[13] performed a combined analysis of both bacterial and fungal communities in cadaver-associated soil beneath two human cadavers using terminal restriction fragment length polymorphism (T-RFLP) on the ITS region. They noted the presence of specific succession patterns during several decomposition stages, although the intrinsic limits of the analysis performed did not allow them to identify specific fungal taxa associated with the observed temporal trend. Chimutsa et al.[21] specifically investigated the fungal community changes in tanks filled with garden soil and a decomposing domestic pig (Sus scrofa) leg positioned at three different depths and at three time points after burial (3, 28 and 77 days). The fungal communities were characterised using polymerase chain reaction and denaturing gradient gel electrophoresis (PCR-DGGE) of the 18S rRNA gene. They did not find any statistically significant shift in fungal community diversity or numerical dominance in either the presence or the absence of the carcass, leading the authors to recommend different approaches such as the use of next-generation sequencing (NGS) to overcome the limitations of their study. Metcalf et al.[14] performed a wider study including both mice (in laboratory settings) and decomposing human bodies placed in outdoor settings at the Sam Houston State University (Southeast Texas Applied Forensic Science Facility). They evaluated at different PMIs the succession of bacteria, archaea and fungi both on skin and within the cadaver-associated soil, highlighting the presence of several patterns characteristic for specific bacterial and fungal succession in the gravesoil analysed at increasing PMIs. Fu et al. performed metabarcoding analyses of the fungal communities associated with the decomposition of rat carcasses[22] and more recently of juvenile pigs both exposed in outdoor conditions (in China, during the summer period, sampling was performed from day 0 to day 14 post-mortem) and left to decompose indoors[23] as a proxy to estimate PMI from the fungal shift. They found trends between specific decomposition stages and specific fungal taxa, suggesting potential for this type of analysis to estimate PMI.

Due to the important effect that environmental variables can have on the decomposition rate[7], it is crucial to perform additional outdoor studies in different geographical locations in order to test the possibility of building a single regression model for PMI estimation from metabarcoding analyses that may be suitable for a wide range of environmental conditions, and to include temperature records in the model to take into account also the seasonality within the same environment[7]. At the best of our knowledge, no such studies involving the metabarcoding of fungal species for forensic purposes have ever been done in the U.K., neither in graves, and this limits the applicability of the results found by other groups in this country. For this reason, we performed an experimental shallow outdoors burial of four pig carcasses at a rural site within the U.K. and investigated PMIs ranging from one to six months. In particular, we collected soil samples from the graves after 1, 2, 4 and 6 months post-mortem and conducted metabarcoding analyses on the samples. Analyses of the bacterial succession through the amplification and sequencing of the 16S rRNA genes were previously performed on soil samples (Procopio et al.[19]). Now our aim is to focus on the fungal community collected from the same gravesoil to obtain a complete profile of the microbial populations associated with specific PMIs. The U.K. is markedly colder than the climate at sites where similar studies have been conducted so far (e.g. Texas or China) and therefore could provide new insights on the estimation of PMI starting from microbial metabarcoding data in colder climates; furthermore, this type of study has never been conducted on grave scenarios before, and for this reason it will provide for the first time an insight on the mycobiome associated with gravesoil for forensic applications.

Another aim of this paper was to elucidate which one of the ITS regions (ITS1 and ITS2) was the most appropriate to be used for DNA metabarcoding studies for forensic applications, since this is a topic of debate in the ecological community. Blaalid et al.[24] obtained similar results when using either ITS1 or ITS2 to perform fungal metabarcoding with environmental data, but Banchi et al.[25] showed discrepancies and biases in the detection of Basidiomycota when ITS1 or ITS2 were used as metabarcode for lichen mycobiomes, and suggested the complementary analysis of both ITS1 and ITS2 to reliably estimate the taxonomic diversity of the samples. In similar studies, Wang et al.[26] concluded conversely that ITS1 provides a better DNA barcode than ITS2 allowing a better species discrimination efficiency, whereas the ITS2 has been specifically selected and preferred to ITS1 in another study to catch fungal diversity in airborne samples[27]. It is clear that an a priori choice of the region/s to target was not ideal, and for this reason, our analysis of the fungal communities targeted specifically both the ITS1 and the ITS2 regions. Aims were both to evaluate if there were differences in terms of their performance in this study, and ultimately to suggest the best option for the forensic community interested in performing similar experiments in other environmental conditions in the future.

At each selected post-mortem interval, the graves were excavated and the carcasses sampled to collect soil samples for the metabarcoding. The decomposition stages and the morphological aspects of the carcasses were evaluated and described in Procopio et al.[19]. Briefly, after one month, P1 was in putrefaction, but overall the carcass shape was still recognizable, after two months, P2 was in the putrefaction/liquefaction stage of decomposition with bones still attached to ligaments, after four months, P3 was partially skeletonised with still some soft tissues present, and after six months, P4 was fully skeletonised and disarticulated.

Metabarcoding analysis of soil fungi

Within the 18 samples analysed, we respectively obtained 1,529,090 sequences after Library A pre-processing and 1,827,776 sequences after Library B pre-processing. Soil controls yielded overall 1,144,877 sequences (from 71,600 to 102,938 for Library A and from 46,536 to 141,094 for Library B), while gravesoil samples yielded overall 2,211,989 sequences (from 60,085 to 113,214 for Library A and from 67,102 to 140,992 sequences for Library B). After the de-novo clustering, sequences with 97% identity were clustered providing 1,640 initial OTUs for Library A and 2,569 initial OTUs for Library B. We identified overall two kingdoms within Library A and 12 within Library B (Supporting Table S1), showing that the ITS2 pair of primers allowed a wider amplification spectrum than the ITS1 pair. OTUs belonging to “Fungi” kingdom were respectively 99.7% in Library A (1,264 out of 1,268) and 81.8% in Library B (1,543 out of 1,885); all the OTUs belonging to different kingdoms were not considered for further analyses. Also, in term of number of reads, Library A showed slightly fewer reads for Fungi (~ 1,500,000) than Library B (~ 1,700,000), as well as fewer OTUs than Library B (Supporting Figure S1). After the filtration step that removed OTUs whose cluster size was smaller than 50 reads (overall 601 OTUs for Library A and 857 OTUs for Library B were withheld after the filtration) and OTUs whose Coefficient of Variation was greater than 3.0, we obtained a final number of 571 OTUs for Library A and 749 OTUs for Library B (Supporting Table S3). The sequence coverage obtained for both libraries showed adequacy for the proposed study (Supporting Figure S2) and that, for both libraries, the control samples were the highest in terms of species richness, followed by P1, and then by P2-P4 that showed similarities to each other.

Fungal community changes during progressive stages of decomposition

Richness and diversity (alpha diversity) of the microbial taxa associated with different PMIs were addressed using the Shannon-Wiener index and boxplots representing the species distribution among samples, with replicates grouped together in terms of raw species number (Fig. 1). Other indices for alpha diversity such as Chao1, abundance-based coverage estimators (ACE), Simpson’s and Fisher’s index have been also applied, as recommended by Bandeira et al.[28], in order to better define the communities (Supporting Figure S3).

The Shannon-Wiener diversity indices (Fig. 1A) showed statistical differences between the mean ranks of at least one pair of groups for Library A and B (Kruskal-Wallis test, p value = 0.027 and 0.012, respectively). However, after Bonferroni correction the only statistical difference found was between controls and P2 for Library B (q value = 0.014). We also calculated the linear regression models for both libraries (Supporting Figure S4). The regression lines calculated for each library did not meet at any point in the graph, meaning that the Shannon-Wiener diversity indices for Library A and B were not statistically different from each other.

Taxonomic richness data were normally distributed (Saphiro-Wilk test, p value = 0.13), and have been consequently evaluated using an analysis of variance (ANOVA) test (which showed a significant change for both libraries, p value < 0.0001) followed by post hoc Tuckey test (Table 1).

Table 1

Post hoc Tuckey test results for taxonomic richness calculated for Library A and B. Non significant values have been reported with the symbol “-“. The symbol “↑” indicates an increase, and the symbol “↓” a decrease.
Library	Control vs P1	P1 vs P2	P2 vs P3	P3 vs P4
A	-	↓ from P1 to P2, p = 0.004	-	-
B	↓ from C1 to P1, p < 0.0001	↓ from P1 to P2, p < 0.0001	-	↑ from P3 to P4, p = 0.003

Library B samples showed smaller standard deviations in the data (Fig. 1B) and allowed for a better discrimination between the various time points than Library A. Results showed an overall decrease in richness of the fungal population with increasing PMIs within both libraries, and this is similar to what we have already found for soil bacteria[19] and to what found by Fu et al.[23] in soil underneath decomposing exposed pigs. This suggests that the presence of a decomposing carcass has a negative impact on the diversity of the fungal communities present in the soil, and the effect is particularly exacerbated when the body reaches the putrefaction/liquefaction stage of decomposition where the least richness has been recorded. The biotic and abiotic factors in the soil environment are altered due to the release of organic compounds in the soil at different decomposition stages[29], and fungal species present in the environment as well as fungi originating from the microflora of the body are subjected to a growth inhibition that can result in sequential changes to the composition of the fungal community[13]. Results obtained here also showed that fungal communities seem to be more sensitive than bacteria to alterations in the soil environment, because the species richness reduction was already significant between C1 and P1, in contrast with what previously observed for bacteria[19]. Due to the dramatic shift in diversity observed with decomposition starting from the first month, it can be argued that similar trends to those observed here will likely be observed regardless of the environment, seasonality, soil type etc., adding strength to the use of fungal metabarcoding for potential PMI estimation and allowing comparisons between different studies. The significant increase in the taxonomic richness of the samples collected after six months PMI showed that fungal communities appear to start returning toward original values faster than the bacterial ones[19]. This could be due to their reduced sensitivity to strict pH parameters in comparison with bacteria[30] and to their potential better adaptation to new environments than bacteria[31]. Despite the quicker return to the original condition for fungi than for bacteria, more than six months appeared to be necessary for both fungi and bacteria to return to their original “basal” state; an observation that may be particularly useful in the investigation of clandestine movements of buried bodies from their original grave to a secondary grave.

To test whether fungal communities were statistically different from each other, we evaluated the homogeneity of dispersion among groups (Fig. 2A) and then run a PERMANOVA test. The test provided significant results (R2 = 0.883 and p = 1e-04 for Library A and R2 = 0.854 and p = 1e-04 for Library B), thus indicating that the dissimilarity was influenced by difference in composition between groups. Consequently, we analysed the differences in the taxonomic abundances from different samples to evaluate the beta diversity between different samples and we used a nonmetric multidimensional scaling ordination (NMDS) plot to visualise Bray Curtis distances (Fig. 2B). Samples collected from the same location clustered better in Library B, with the only exceptions being for P1 and P4, where the clusters were less scattered in Library A. Overall, P1 was closer to the control, as previously reported also for bacterial data[19], whereas samples collected with increasing PMIs showed an ordered distribution from the control from the left to the right on the plot, with P2 being more distant, P3 being the most distant one and P4 being closer again to the control. Overall results suggest that ITS1 and ITS2 performed similarly in the characterisation of the various samples, despite ITS2 was more accurate than ITS1.

To give a comparison of the NMDS ordinations of Bray-Curtis distances for Library A and B, we performed the Procrustes analysis that provides a plot where the same samples are connected by vector arrows. Results showed that the degree of concordance between Library A and B is statistically significant (p = 1e-04), meaning that the ordination of the data is similar between the two libraries (Supporting Figure S5).

To characterise the differences found between the samples, we looked at the abundance of phyla (Supporting Figure S6). Due to the incompleteness of the database for fungal species, we were not able to identify all of the species sequenced, and so we had to accept some unidentified phyla in the data.

When looking at specific PMIs, there were some noticeable trends with increasing time elapsed since death at phyla (Fig. 3) and family level (Fig. 4, Table 2 and Supporting Table S4). In particular, Ascomycota showed a notable increase in abundance after two-months and a decrease after four- and six-months post-mortem compared with the controls for both libraries, despite their richness decreasing after two-months and slowly starting to increase again from four-months onward. Abundance of Pyronemataceae increased significantly in P2 in comparison with the control, and decreased after four months post-mortem. Abundance of Melanommataceae, Bionectriaceae, Massarinaceae, Onygenaceae Pseudeurotiaceae, Coniochaetaceae, Gymnoascaceae, Chaetomiaceae and Nectriaceae on the other side showed a decline in P2, P3 and P4 compared with the control, indicating that the gravesoil environmental composition negatively affected their survival. Ascomycota are decomposers able to break down the animal tissues into major organic compounds, and they have already been identified as fungal decomposers particularly active during the active decay of the carcass[32]. The increase of the abundance of Ascomycota during the first stages of decomposition was not reported by Fu et al.[23] within soil collected underneath exposed pig cadavers, and may be specific to buried cadavers.

Basidiomycota showed a constant decrease in their abundance starting from a higher amount and variety of families in control samples and ending in a significantly lower amount with prolonged PMIs. This result is in contrast with that of Metcalf et al.[32], who suggested that Basidiomycota are particularly active during the active decay stage, but is in agreement with Fu et al.[33], who noticed a decrease in the abundance of this phylum in the rectum of decomposing rats with increasing time after death. In our case, the carcasses reached the active stage of decomposition between one and two months, but Basidiomycota decreased rapidly with the introduction of the carcass in the soil (from control to P1), with Piskurozymaceae family being the most abundant one after one month, and with all the families, including Entolomataceae, Psathyrellaceae disappearing completely after two, four- and six-months post-mortem, suggesting their limited role during the active decomposition of the carcasses in this study. The same happened for the Glomeromycota phylum, where Diversisporaceae constantly decreased from P2 onwards.

Mortierellomycota, and in particular Mortierellaceae, increased from the control to P1, then decreased in P2, and further increased in P3 and P4. This is in agreement with what was found also by Metcalf et al.[32], namely that Mortierellaceae are active particularly in between the active decay and the dry remains stages. However, we also noticed their increase in P1 in comparison with the control, suggesting that the Mortierellomycota also may have a role during the early decomposition stages of the carcasses, although this could be minor compared with their activity during the advanced stages of decay. We did not find an increase in Zygomycota and in Chytridiomycota with time, as found in contrast by Fu et al.[33] in the rectum of decomposing rats.

Overall, the results obtained here indicated the presence of specific patterns of increase or decrease of particular families with advancing decomposition stages; results found in Library A and B were sometimes specific for the specific library investigated (e.g., some families were found statistically significant only for one of the two libraries), but the trends found were always in agreement between the two (e.g., Massarinaceae decreased constantly and significantly both for Library A and B). Unfortunately, the very limited amount of metabarcoding data available on fungal communities associated with the cadaveric decomposition did not allow us to perform a comprehensive comparison between different studies. However, we found some similarities both with the study conducted by Metcalf et al.[32] on human cadavers in Texas, with the one conducted by Fu et al.[44] on rat carcasses in China and with the following study conducted by the same group on exposed pigs[23], despite in all cases, they were conducted with exposed remains instead of with buried ones as in our study. Overall we can state that the fungal succession during the various decomposition stages appears reproducible and reliable despite different geographical areas, different animal carcasses and different post-mortem conditions (e.g. buried vs non-buried). This shows the potential for this type of analyses to provide a fungal “clock of death” similar to the one already proposed for bacteria, that will be applicable to worldwide contexts, provided that more data should be collected from other experiments conducted in different conditions and in different geographical areas.

Table 2

Families grouped according to their phylum showing statistical differences between the control and the various time points (P1 to P4). Statistical significance was calculated via Student’s t-test with False Discovery Rate (FDR) correction. Results reported here were selected for p values ≤ 0.005. The symbol “↑” indicates an increase in comparison to the control, and the symbol “↓” a decrease. Non significant values have been reported with the symbol “-“.
	Library A				Library B
	P1	P2	P3	P4	P1	P2	P3	P4
	Ascomycota
Massarinaceae	↓ p = 0.004	↓ p = 0.0006	↓ p = 0.0007	↓ p = 0.001	↓ p = 0.002	↓ p = 0.0009	↓ p = 0.0008	↓ p = 0.005
Massarinaceae	↓ p = 0.004	↓ p = 0.0006	↓ p = 0.0007	↓ p = 0.001	↓ p = 0.002	↓ p = 0.0009	↓ p = 0.0008	↓ p = 0.005
Pyronemataceae	-	↑ p = 0.002	-	-	-	↑ p = 0.003	-	-
Pyronemataceae	-	↑ p = 0.002	-	-	-	↑ p = 0.003	-	-
Melanommataceae	-	↓ p = 0.004	↓ p = 0.004	↓ p = 0.004	-	-	-	-
Melanommataceae	-	↓ p = 0.004	↓ p = 0.004	↓ p = 0.004	-	-	-	-
Bionectriaceae	-	↓ p = 0.002	↓ p = 0.003	↓ p = 0.002	-	-	-	-
Bionectriaceae	-	↓ p = 0.002	↓ p = 0.003	↓ p = 0.002	-	-	-	-
Onygenaceae	-	↓ p = 0.002	-	-	-	-	-	-
Onygenaceae	-	↓ p = 0.002	-	-	-	-	-	-
Pseudeurotiaceae	-	-	-	-	-	↓ p = 0.004	↓ p = 0.005	-
Pseudeurotiaceae	-	-	-	-	-	↓ p = 0.004	↓ p = 0.005	-
Coniochaetaceae	-	-	-	-	-	↓ p = 0.003	↓ p = 0.003	↓ p = 0.005
Coniochaetaceae	-	-	-	-	-	↓ p = 0.003	↓ p = 0.003	↓ p = 0.005
Gymnoascaceae	-	-	-	-	-	↓ p = 0.004	-	-
Gymnoascaceae	-	-	-	-	-	↓ p = 0.004	-	-
Chaetomiaceae	-	-	-	-	-	↓ p = 0.003	-	↓ p = 0.004
Chaetomiaceae	-	-	-	-	-	↓ p = 0.003	-	↓ p = 0.004
	Zoopagomycota
Piptocephalidaceae	↑ p = 0.004	-	-	-	-	-	-	-
Piptocephalidaceae	↑ p = 0.004	-	-	-	-	-	-	-
	Zygomycota
Mortierellaceae	-	-	-	↑ p = 0.003	-	-	-	-
Mortierellaceae	-	-	-	↑ p = 0.003	-	-	-	-
	Basidiomycota
Entolomataceae	-	-	-	-	-	↓ p = 0.005	-	-
Entolomataceae	-	-	-	-	-	↓ p = 0.005	-	-
Psathyrellaceae	-	-	-	-	-	↓ p = 0.0009	↓ p = 0.0008	↓ p = 0.004
Psathyrellaceae	-	-	-	-	-	↓ p = 0.0009	↓ p = 0.0008	↓ p = 0.004
	Glomeromycota
Diversisporaceae	-	-	-	-	-	↓ p = 0.005	↓ p = 0.005	↓ p = 0.005
Diversisporaceae	-	-	-	-	-	↓ p = 0.005	↓ p = 0.005	↓ p = 0.005

Identification of endogenous mammalian fungal communities and basal communities

In order to identify specifically basal fungal communities that vary depending on the PMI of the bodies, excluding the taxa introduced in the cadaveric soil by the carcass (e.g. skin fungi present in and on the bodies), we evaluated which taxa were uniquely present in the cadaveric soil and not in the control samples, and then excluded these unique OTUs from the analysis of the “basal community”. In summary, for Library A we found 569 different OTUs present within the basal microbiome (Supporting Table S5) and only 2 taxa specifically associated with the bodies (not identified in the control, Supporting Table S6), and for Library B we found 735 different OTUs present within the basal microbiome (Supporting Table S5) and 14 taxa specifically associated with the cadaveric soil only (not identified in the control) (Supporting Table S6). It can be noted here that primers used for Library B allowed a better identification of fungi associated with the decomposing carcass than primers for Library A. For this reason, we will discuss results regarding the mycobiome introduced by the carcass only with reference to Library B (Fig. 5). The results showed that the carcass’ fungal community that was abundant in soil at the first time point, almost disappeared with advanced decomposition stages, apart from some outliers such as the Onygenales order (family “Incertae sedis”, genus Chrysosporium) which were more abundant in P1 and P3, and the Psathyrellaceae (genus Coprinellus) which started to increase only after six months. The term “Incertae sedis” found in Fig. 6A is a mycological term used to combine together different taxa which have not been classified yet into a specific family, and that for this reason can only be grouped in this way.

These findings could suggest that the cadaveric soil is not a favourable environment for the survival and for the growth of the fungi introduced with the carcass, and that overall the fungi introduced into the soil by the body are replaced by soil fungi in less than two months. This finding is also in accordance with what found by Fu et al.[23], where the variety of soil fungi present after one-day post-mortem was higher than that found during the decomposition of the carcasses. In particular, Cladosporiaceae family has been found to be a physiologic inhabitant of the intestine of the pig[34], and our results showed that Cladosporiaceae spp. are not able to survive with advancing decomposition stages. However, we did not find evidence in living pigs for the other families that were found in abundance in P1, raising a question around the decrease, and then the eventual increase after several months, of some of these fungal families in the cadaveric soils. Interestingly, members of the Mortierellaceae family were found more abundant in P1 and then not present at prolonged PMIs; altough this may seem to contradict the previous findings, it should be noted that the specific taxa mentioned in this section of the work are the ones not found in the control soil and exclusively brought in by the carcass itself. So, despite some taxa of this family not being able to survive after prolonged PMIs, other members of the Mortierellaceae family already present in the soil showed the ability to take advantage of the organic nutrients released with the decomposition of the carcasses to become the most abundant species in the cadaberic soil after four- and six-months post mortem.

Indicator species analysis was also undertaken to evaluate which species were significantly associated specifically with the basal microbiome or with the carcasses (significance level = 5%) (Supporting Table S7). Overall, the analysis showed that 253 OTUs for Library A and 295 species for Library B contributed to explain the differences observed between the control and the gravesoil (p values ≤ 0.05), and that two OTUs contributed significantly to discriminate carrion-associated fungi from the basal fungal communities for Library B, but none for Library A. This result further proves how limited the contribution to the soil fungal communities introduced by the carrion is, and supports the potential to extend these findings to future applications involving the presence of human carrions.

Due to the fact that the fungal species associated with specific PMIs predominantly originated from the soil and not from the pig carcasses, we could suggest the use of results found in Table 2 as good indicators or biomarkers for specific PMIs. To further prove this, we performed a Student’s t-test with FDR correction to report only the taxa associated with significant increases or decreases at selected PMIs. Results found were almost identical to the ones found in Table 2, with the same families associated with significant differences between the control and the various PMIs (Supporting Figures S7-10 and Supporting Table S8). This result further proves the potential of these results to be applied also to forensic cases or studies in a similar geographical area but involving the presence of decomposing humans instead of animal models such as pigs.

This study showed that shifts in the fungal communities present in the gravesoil, like bacterial ones, have the potential to reveal information about the PMI of buried carcasses from one- to six-months post mortem. Here we were able to demonstrate that specific fungal families, such as the Mortierellaceae or the Massarinaceae, may be ideal for the estimation of the PMI; providing that additional research with additional bodies and time points is done to improve the accuracy of the findings presented here. According to our results, the decrease of Masssarinaceae families in cadaveric soil compared to a control soil may indicate the presence of a decomposing body, and the reduction in abundance of specific Ascomycota families such as Melanommataceae, Bionectriaceae, Pseudeurotiaceae, Coniochaetaceae and Chaetomiaceae, Basidiomycota families such as Psathyrellaceae and Glomeromycota families such as Diversisporaceae may indicate the presence of a body in active or advanced decomposition stages. On the contrary, the presence of an increased abundance of Pyronemataceae in the soil, in comparison with a control, may indicate the presence of a body in the active decomposition stage. An increase of Piptocephalidaceae might indicate the presence of a body at the beginning of its decomposition, and the increase of Mortierellaceae might indicate the presence of a body in an advanced decomposition stage. We have also showed here that the specific analysis of the ITS1 or the ITS2 sites provides similar results, but that the range of data obtained with ITS2 is generally wider when compared to ITS1, allowing for a better characterisation of the distinct PMIs. For this reason, we recommend here either the combination of the two ITS sites, or to exclusively target the ITS2 region in future instead of using the ITS1 only. The limited availability of metabarcoding fungal data in forensic caseworks, combined with the lack of curated databases, partially reduced the possibility of fully understanding the reasons behind the findings and required the manual editing of the data by a mycologist. Although this could limit the applications of these type of analyses for forensic applications, we are confident that this work will provide useful insights on the fungal shifts in the burial environment in temperate climates, and might be the starting point for additional research in this field.

Experimental design

The experimental design of the study has been already described in detail in the first part of this work, which focused on the analysis of the bacterial communities within the burial context[19], and is summarised here. Four juvenile pigs (Sus scrofa) (named P1, P2, P3 and P4) were experimentally buried under approximately 40 cm soil at the HuddersFIELD outdoor taphonomy facility (University of Huddersfield, U.K.) between May and November 2016. The experiment was conducted following the guidelines of the U.K. Department for Environment, Food and Rural Affairs (DEFRA). Soil physicochemical and biological characteristics recorded before the burials are reported in Procopio et al.[19]. Pigs of similar age (3–5 weeks old) and weight (4.5–11 kg) which had died from natural causes were kept frozen after death until their burial. The experiment time was set to zero when the cadavers were allowed to de-frost and were placed within soil graves, and subsequently each pig was left in the grave for a specific amount of time post-mortem (P1 = one month, P2 = two months, P3 = four months, P4 = six months). Soil samples in contact with the superior surface of the carcass (from the fore-limb area, the abdominal area and the hind-limb area of the body) were pooled in one bag per individual/burial at the selected PMI, and frozen at -20 °C after their collection until further analysis. Three technical replicates per bag (i.e., one bag per grave) were subsampled from each bag and used for metabarcoding analyses. Additionally, control soil samples (taken at similar soil depths but without a carcass being present) were collected at locations two and six meters away from the graves. The control samples were obtained concurrently with the last sampling performed in November 2016 (six months PMI), to provide a representative record of the background microbial activity within the field used for the experiment without direct microbial contamination from the graves. According to Metcalf et al.[14], the observable differences within the microbial community associated with the close proximity of a decomposing body are larger and more significant than those caused by variations in local environmental conditions (linked with seasonality, temperature and rainfall). For this reason, our control samples were collected at different distances from the graves but at a single time point. Control soil samples were frozen at -20 °C until further analysis. Metabarcoding samples totalled 18 (three samples per grave plus six controls). Soil pHs were recorded and reported in Procopio et al.[19], as well as soil temperatures and local temperatures and rainfalls (Procopio et al.[35]).

DNA extraction, amplification and sequencing

Genomic DNA was extracted from 500 mg of soil sample using the FastDNA→ SPIN Kit for Soil (MP Biomedicals, Europe), following the guidelines provided by the manufacturer, and fungal communities were specifically targeted amplifying the internal transcribed spacer (ITS) region in the nuclear ribosomal repeat unit ITS1 and ITS2. Forward ITS1-F (CTTGGTCATTTAGAGGAAGTAA) and reverse ITS2 (GCTGCGTTCTTCATCGATGC) primers for ITS1 [36] and forward fITS9 (GAACGCAGCRAAIIGYGA) and reverse ITS4 (TCCTCCGCTTATTGATATGC) primers (Sigma-Aldrich, U.K.) for ITS2 ([37],[38] respectively) were used in combination with 8 bp unique tags (Supporting Table S2). The expected amplicon size was ~ 250–600 bp for ITS1[39, 40] and ~ 240–460 for ITS2. Primers and tags used for the subsequent analyses both for ITS1 and for ITS2 amplifications have been listed in Supporting Table S2. Degeneracies were allowed in the primers to decrease eventual biases and to increase the ability to detect organisms characterised by slight modifications in their binding site’s sequences. PCR negative controls were run in each analysis to perform a quality check of the amplifications and in all cases these controls gave negative results. PCR positive controls (Saccharomyces cerevisiae purified DNA) always showed a good signal in the gel, confirming the success of each PCR performed. PCR reaction mixtures were set up as follows: 12.5 µL master mix (Platinum Hot Start PCR Master Mix 2x, Thermo Fisher U.K.), 0.5 µL forward primer (10 µM), 0.5 µL reverse primer (10 µM) and 0.5–1.5 µL template DNA in a final reaction volume of 25 µL. The thermal cycler (T3000, Biometra GmbH, Göttingen, Germany) conditions were set up as follows: denaturation at 94 °C for 2 min; 35 cycles of denaturation at 94 °C for 30 sec, annealing at 52 °C for 40 sec and extension at 68 °C for 30 sec; final extension at 68 °C for 10 min and maintenance of the samples at 4 °C. The PCR products obtained were checked on 1.5% (w/v) agarose gels (Sigma-Aldrich, U.K.), purified (Wizard® SV Gel and PCR Clean-UpSystem, Promega, U.K.), quantified with Qubit (Qubit Fluorometric Quantitation, Thermo Fisher Scientific, U.K.) and sent to IGA Technology Services (Udine, Italy) for paired-end sequencing using the Illumina MiSeq technology (2 × 250 bp). The company ligated our tagged PCR products with a couple of Illumina adapters, and generated two libraries named “Library A” and “Library B” for ITS1 and ITS2, respectively, due to the different length of the amplicons expected for ITS1 and ITS2 as previously mentioned. Raw data for Library A and B were deposited in the NCBI Sequence Read Archive (SRA-NCBI; https://www.ncbi.nlm.nih.gov/sra) under project accession number PRJNA594906 and BioSample accession numbers SAMN13816975-SAMN13816980.

DNA data analysis

Paired-end reads from each library were merged using Pear v.0.9.2[41], with the quality score threshold for trimming the low‐quality part of a read set at 28, the minimum length of reads after the trimming process set at 200 bp and the minimum possible length of the assembled sequences set at 200 bp. Unix bash commands were used to select assembled sequences beginning with recognisable forward primer (no mismatch allowed), to trim initial and terminal 19 and 20 bases (corresponding to forward and reverse primers respectively) and to assign a sample specific progressive count to each fragment. Assembled sequences from each library were clustered into Operational Taxonomic Units (OTUs) using a de novo clustering strategy with QIIME v1.91[42] and VSEARCH v2.3.4[43] (https://github.com/torognes/vsearch) at 97% similarity. Taxonomic assignment to OTUs was performed using the full "UNITE + INSD" dataset v.8.2 (released 2019-02-02; [44]) as a reference, and using BLAST and UCLUST, implemented in QIIME, as assignment methods. A consensus of the two assignment methods has been manually inspected and edited by mycologists and used for subsequent statistical analyses.

Statistical Analyses

Numerical ecology statistical analyses were performed within the computing environment R (https://www.R-project.org/)[45]. As previously described in Procopio et al.[19], control samples collected from two different locations were extracted in triplicate (replicate #1, #2 and #3) from each location (overall six DNA samples extracted), and during the data analysis they have been unified arbitrarily into a single control sample, summing up the reads of each of the three replicates to obtain our three replicates for control samples.

A number of filtering steps were applied to gather the final OTU dataset: OTUs with fewer than 50 reads were filtered out as well as samples with fewer than 20 reads and OTUs showing a Coefficient of Variation greater than 3.0 were also removed.

In order to standardize sampling efforts and to allow for comparisons of samples with non-uniform coverage, the OTU table has been normalized by subsampling at even sequencing depth from each sample using the minimal number of sequences obtained between the two libraries as value for the rarefaction (59,413 sequences) by means of the rarefy_even_depth function in the R package phyloseq v.1.22.3[46]. The rarefaction effects on an OTU table are commonly rendered by means of the rarefaction curves, in order to graphically estimate species richness. Raw species richness counts can only be compared when the species richness has reached a clear asymptote; all the species present in a sample are well described when the curve ascribed to each sample reach its plateau. Rarefaction curves were rendered by means of the function ggrare, provided by the richness.R script from the phyloseq extension package by Mahendra Mariadassou (https://github.com/mahendra-mariadassou/phyloseq-extended). All taxon abundances were calculated and graphically plotted with the aim of the R package phyloseq[46].

Biodiversity analyses were carried out by comparing the richness (number of species) and evenness (richness taking into account relative abundances) of microbial communities of the PMIs. Within-sample (alpha) diversity was assessed by six estimators: "observed number of species", "Chao1", "ACE", "Shannon", "Simpson", "Fisher". The alpha diversity indices were calculated and plotted by means of the functions estimate_richness and plot_richness implemented in the R package phyloseq[46].

In order to test whether communities were statistically different from each other, a multivariate homogeneity of group dispersions among the different Control and PMI groups was first assessed by means of the betadisper and permutest (with 9,999 permutations) functions in the R package vegan v.2.5.2[47]. The differences in the composition of fungal communities in PMIs for both libraries were rendered by means of a NMDS, to visualise Bray-Curtis distances, using the functions vegdist and metaMDS in the R package vegan[47].

The permutational multivariate analysis of variance (PERMANOVA, 9,999 permutations), as implemented in the adonis function of the vegan package of R[47], were applied to access whether communities were statistically different from each other.

Indicator species analysis (a classification-based method to measure associations between species and groups of sites[48]) was carried out using the multipatt function in the indicspecies v.1.7.6 R package, with 9,999 permutations[49] in order to assess whether OTUs (and if so, which ones) were significantly associated with a particular PMI.

To explore the shifts in the fungal basal communities at different decomposition stages, we evaluated which families showed statistically significant differences in the alpha diversity using the Kruskal-Wallis test and the Dunn’s pairwise tests, and in beta diversity and in the abundances at different PMIs applying the Student’s t-test with Benjamini and Hochberg False Discovery Rate (FDR) correction[50] with the T.TEST and P.ADJUST functions in R. This generated both p values and adjusted p values (q values), that were used to evaluate the statistical significance of the results reported in the work.

List of abbreviations

abundance-based coverage estimators (ACE)

analysis of variance (ANOVA)

Department for Environment, Food and Rural Affairs (DEFRA)

False Discovery Rate (FDR)

Internal Transcribed Spacer (ITS)

next-generation sequencing (NGS)

nonmetric multidimensional scaling ordination (NMDS)

operational taxonomic units (OTUs)

polymerase chain reaction (PCR)

polymerase chain reaction and denaturing gradient gel electrophoresis (PCR-DGGE)

permutational multivariate analysis of variance (PERMANOVA)

post-mortem interval (PMI)

restriction fragment length polymorphism (T-RFLP)

List of abbreviations

abundance-based coverage estimators (ACE)

analysis of variance (ANOVA)

Department for Environment, Food and Rural Affairs (DEFRA)

False Discovery Rate (FDR)

Internal Transcribed Spacer (ITS)

next-generation sequencing (NGS)

nonmetric multidimensional scaling ordination (NMDS)

operational taxonomic units (OTUs)

polymerase chain reaction (PCR)

polymerase chain reaction and denaturing gradient gel electrophoresis (PCR-DGGE)

permutational multivariate analysis of variance (PERMANOVA)

post-mortem interval (PMI)

restriction fragment length polymorphism (T-RFLP)

Ethics approval: Animals used for the study died for natural causes. The experiment was conducted following the guidelines of the U.K. Department for Environment, Food and Rural Affairs (DEFRA).

Consent for publication: Not applicable

Availability of data and material: The datasets generated and/or analysed during the current study are available in the NCBI Sequence Read Archive (SRA-NCBI; https://www.ncbi.nlm.nih.gov/sra) repository under project accession number PRJNA594906 and BioSample accession numbers SAMN13816975-SAMN13816980.

Competing interests: The authors declare that they have no competing interests.

Funding: The authors are grateful to the Royal Society for funding both a Ph.D. studentship (N.P.) and University Research Fellowship (M.B.) under grants RG130453 and UF120473, respectively. Funding was used to purchase the material required and to perform the analyses. The funding body did not have roles in the design of the study, in the collection, analysis and interpretation of data, neither in writing the manuscript.

Authors contribution: N.P., A.M. and M.B. designed the study. N.P., A.C., A.W. and M.B. performed the experimental burial and the soil samplings. N.P. and A.M. performed the extraction and amplification of DNA. S.G., S.V. and M.C. performed the bioinformatic analysis of the results. N.P., S.G. and S.V. performed the data interpretation. N.P. was the major contributor in writing the manuscript. A.C., M.B. and A.W. performed the proofreading of the manuscript. All authors read and approved the final manuscript.

Acknowledgements: Not applicable.

Zhou W, Bian Y. Thanatomicrobiome composition profiling as a tool for forensic investigation. Forensic Sci Res. 2018;3:105–10. doi:10.1080/20961790.2018.1466430.
Orgiazzi A, Bianciotto V, Bonfante P, Daghino S, Ghignone S, Lazzari A, et al. 454 Pyrosequencing Analysis of Fungal Assemblages from Geographically Distant, Disparate Soils Reveals Spatial Patterning and a Core Mycobiome. Diversity. 2013;5:73–98. doi:10.3390/d5010073.
Orgiazzi A, Dunbar MB, Panagos P, de Groot GA, Lemanceau P. Soil biodiversity and DNA barcodes: opportunities and challenges. Soil Biol Biochem. 2015;80:244–50. doi:10.1016/J.SOILBIO.2014.10.014.
Klindworth A, Pruesse E, Schweer T, Peplies J, Quast C, Horn M, et al. Evaluation of general 16S ribosomal RNA gene PCR primers for classical and next-generation sequencing-based diversity studies. Nucleic Acids Res. 2013;41:e1. doi:10.1093/nar/gks808.
Schoch CL, Seifert KA, Huhndorf S, Robert V, Spouge JL, Levesque CA, et al. Nuclear ribosomal internal transcribed spacer (ITS) region as a universal DNA barcode marker for Fungi. Proc Natl Acad Sci. 2012;109:6241–6. doi:10.1073/pnas.1117018109.
Taberlet P, Prud’Homme S, Campione E, Roy J, Miquel C, Shehzad W, et al. Soil sampling and isolation of extracellular DNA from large amount of starting material suitable for metabarcoding studies. Mol Ecol. 2012;21:1816–20. doi:10.1111/j.1365-294X.2011.05317.x.
Metcalf JL. Estimating the postmortem interval using microbes: Knowledge gaps and a path to technology adoption. Forensic Sci Int Genet. 2019;38:211–8. doi:10.1016/J.FSIGEN.2018.11.004.
Hawksworth DL, Wiltshire P. Forensic mycology: current perspectives. Res Reports Forensic Med Sci. 2015;5:75. doi:10.2147/RRFMS.S83169.
Bellini E, Ambrosio E, Zotti M, Nucci G, Gabbrielli M, Vanezis P. The Usefulness of Cadaveric Fungi as an Investigation Tool. Am J Forensic Med Pathol. 2016;37:23. doi:10.1097/PAF.0000000000000210.
M. Christensen A, Passalacqua N, Bartelink E. Forensic Taphonomy. In: Forensic Anthropology Current Methods and Practice. 1st edition. Academic Press; 2014. p. 119–47.
Swift B. The Timing of Death. In: Rutty GN, editor. Essentials of Autopsy Practice. London: Springer London; 2006. p. 189–214. doi:10.1007/b136465.
Szelecz I, Lösch S, Seppey CVW, Lara E, Singer D, Sorge F, et al. Comparative analysis of bones, mites, soil chemistry, nematodes and soil micro-eukaryotes from a suspected homicide to estimate the post-mortem interval. Sci Rep. 2018;8:25. doi:10.1038/s41598-017-18179-z.
Parkinson RA, Dias K-R, Horswell J, Greenwood P, Banning N, Tibbett M, et al. Microbial Community Analysis of Human Decomposition on Soil. In: Ritz K, Dawson L, Miller D, editors. Criminal and Environmental Soil Forensics. Springer; 2009. p. 379–94.
Metcalf JL, Xu ZZ, Weiss S, Lax S, Van Treuren W, Hyde ER, et al. Microbial community assembly and metabolic function during mammalian corpse decomposition. Science (80- ). 2016;351:158–62.
Finley SJ, Pechal JL, Benbow ME, Robertson BK, Javan GT. Microbial signatures of cadaver gravesoil during decomposition. Microb Ecol. 2016;71:524–9.
Cobaugh KL, Schaeffer SM, DeBruyn JM. Functional and structural succession of soil microbial communities below decomposing human cadavers. PLoS One. 2015;10:e0130201.
Pechal JL, Crippen TL, Tarone AM, Lewis AJ, Tomberlin JK, Benbow ME. Microbial community functional change during vertebrate carrion decomposition. PLoS One. 2013;8:e79035.
Guo J, Fu X, Liao H, Hu Z, Long L, Yan W, et al. Potential use of bacterial community succession for estimating post-mortem interval as revealed by high-throughput sequencing. Sci Rep. 2016;6:24197. doi:10.1038/srep24197.
Procopio N, Ghignone S, Williams A, Chamberlain A, Mello A, Buckley M. Metabarcoding to Investigate Changes in Soil Microbial Communities within Forensic Burial Contexts. Forensic Sci Int Genet. 2018;0. doi:10.1016/j.fsigen.2018.12.002.
Metcalf JL, Parfrey LW, Gonzalez A, Lauber CL, Knights D, Ackermann G, et al. A microbial clock provides an accurate estimate of the postmortem interval in a mouse model system. Elife. 2013;2:e01104.
Chimutsa M, Olakanye AO, Thompson TJU, Ralebitso-Senior TK. Soil fungal community shift evaluation as a potential cadaver decomposition indicator. Forensic Sci Int. 2015;257:155–9.
Fu XL, Guo JJ, Zhu ZY, Ding ZY, Zha L, Cai JF. The potential use of fungi community in postmortem interval estimation in China. Forensic Sci Int Genet Suppl Ser. 2015;5:e476–8. doi:10.1016/J.FSIGSS.2015.09.189.
Fu X, Guo J, Finkelbergs D, He J, Zha L, Guo Y, et al. Fungal succession during mammalian cadaver decomposition and potential forensic implications. Sci Rep. 2019;9.
Blaalid R, Kumar S, Nilsson RH, Abarenkov K, Kirk PM, Kauserud H. ITS1 versus ITS2 as DNA metabarcodes for fungi. Mol Ecol Resour. 2013;13:218–24. doi:10.1111/1755-0998.12065.
Banchi E, Stankovic D, Fernández-Mendoza F, Gionechetti F, Pallavicini A, Muggia L. ITS2 metabarcoding analysis complements lichen mycobiome diversity data. Mycol Prog. 2018;17:1049–66. doi:10.1007/s11557-018-1415-4.
Wang X-C, Liu C, Huang L, Bengtsson-Palme J, Chen H, Zhang J-H, et al. ITS1: a DNA barcode better than ITS2 in eukaryotes? Mol Ecol Resour. 2015;15:573–86. doi:10.1111/1755-0998.12325.
Banchi E, Ametrano CG, Stanković D, Verardo P, Moretti O, Gabrielli F, et al. DNA metabarcoding uncovers fungal diversity of mixed airborne samples in Italy. PLoS One. 2018;13:e0194489. doi:10.1371/journal.pone.0194489.
Bandeira B, Jamet J-L, Jamet D, Ginoux J-M. Mathematical convergences of biodiversity indices. Ecol Indic. 2013;29:522–8.
Vass AA, Smith RR, Thompson C V, Burnett MN, Wolf DA, Synstelien JA, et al. Decompositional odor analysis database. J Forensic Sci. 2004;49:760–9. http://www.ncbi.nlm.nih.gov/pubmed/15317191. Accessed 8 Jun 2019.
Finley SJ, Benbow ME, Javan GT. Potential applications of soil microbial ecology and next-generation sequencing in criminal investigations. Applied Soil Ecology. 2015;88:69–78. doi:10.1016/j.apsoil.2015.01.001.
Selbmann L, Egidi E, Isola D, Onofri S, Zucconi L, de Hoog GS, et al. Biodiversity, evolution and adaptation of fungi in extreme environments. Plant Biosyst - An Int J Deal with all Asp Plant Biol. 2013;147:237–46. doi:10.1080/11263504.2012.753134.
Metcalf JL, Carter DO, Knight R. Microbiology of death. Curr Biol Mag. 2016;26:R543–76. doi:10.1016/j.cub.2016.03.042.
Fu XL, Guo JJ, Zhu ZY, Ding ZY, Zha L, Cai JF. The potential use of fungi community in postmortem interval estimation in China. Forensic Sci Int Genet Suppl Ser. 2015.
Hurst CJ. Advances in Environmental Microbiology. Springer; 2016.
Procopio N, Williams A, Chamberlain AT, Buckley M. Forensic proteomics for the evaluation of the post-mortem decay in bones. J Proteomics. 2018;177:21–30.
Gardes M, Bruns TD. ITS primers with enhanced specificity for basidiomycetes - application to the identification of mycorrhizae and rusts. Mol Ecol. 1993;2:113–8. doi:10.1111/j.1365-294X.1993.tb00005.x.
Ihrmark K, Bödeker ITM, Cruz-Martinez K, Friberg H, Kubartova A, Schenck J, et al. New primers to amplify the fungal ITS2 region - evaluation by 454-sequencing of artificial and natural communities. FEMS Microbiol Ecol. 2012;82:666–77. doi:10.1111/j.1574-6941.2012.01437.x.
White TJ, Bruns T, Lee S, Taylor J. Amplification and Direct Sequencing of Fungal Ribosomal RNA Genes for Phylogenetics. PCR Protoc. 1990;:315–22. doi:10.1016/B978-0-12-372180-8.50042-1.
Bokulich NA, Mills DA. Improved selection of internal transcribed spacer-specific primers enables quantitative, ultra-high-throughput profiling of fungal communities. Appl Environ Microbiol. 2013;79:2519–26.
Hoggard M, Vesty A, Wong G, Montgomery JM, Fourie C, Douglas RG, et al. Characterizing the human mycobiota: A comparison of small subunit rRNA, ITS1, ITS2, and large subunit rRNA genomic targets. Front Microbiol. 2018;9 SEP.
Zhang J, Kobert K, Flouri T, Stamatakis A. PEAR: a fast and accurate Illumina Paired-End reAd mergeR. Bioinformatics. 2013;30:614–20.
Caporaso JG, Kuczynski J, Stombaugh J, Bittinger K, Bushman FD, Costello EK, et al. QIIME allows analysis of high-throughput community sequencing data. Nat Methods. 2010;7:335–6. doi:10.1038/nmeth.f.303.
Rognes T, Flouri T, Nichols B, Quince C, Mahé F. VSEARCH: a versatile open source tool for metagenomics. PeerJ. 2016;4:e2584.
Nilsson RH, Larsson K-H, Taylor AFS, Bengtsson-Palme J, Jeppesen TS, Schigel D, et al. The UNITE database for molecular identification of fungi: handling dark taxa and parallel taxonomic classifications. Nucleic Acids Res. 2019;47:D259–64. doi:10.1093/nar/gky1022.
Team RC. R: A language and environment for statistical computing. 2013.
McMurdie PJ, Holmes S. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS One. 2013;8:e61217.
Oksanen J, Blanchet FG, Friendly M, Kindt R, Legendre P, McGlinn D, et al. Vegan: Community Ecology Package. R package version 2.4-6. 2018.
Dufrêne M, Legendre P. Species assemblages and indicator species: the need for a flexible asymmetrical approach. Ecol Monogr. 1997;67:345–66.
Cáceres M De, Legendre P. Associations between species and groups of sites: indices and statistical inference. Ecology. 2009;90:3566–74.
Benjamini Y, Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J R Stat Soc Ser B. 1995;:289–300.

Download PDF

Journal Publication

published 24 Jul, 2020

Read the published version in Frontiers in Microbiology →

Version 1

posted

You are reading this latest preprint version

Soil Fungal Communities Investigated by Metabarcoding within Simulated Forensic Burial Contexts

Status:

Journal Publication

Version 1

Abstract

Figures

Background

Results And Discussion

Conclusions

Methods

Declarations

References

Supplementary Files

Status:

Journal Publication

Version 1