{"paper_id":"492a39e9-8fd0-40e1-b128-644c7033fb3c","body_text":"PREPRINT\nAuthor-formatted, not peer-reviewed document posted on 01/02/2023\nDOI: https://doi.org/10.3897/arphapreprints.e101101\nAssessing the diversity and spatial distribution of\nnematodes in the Store Mosse National Park (Sweden)\nusing metabarcoding\nMohammed Ahmed, Dieter Slos,  Oleksandr Holovachov\n\n \nAssessing the diversity and spatial distribution of nematodes in the \nStore Mosse National Park (Sweden) using metabarcoding \n \n \nMohammed Ahmed1, Dieter Slos2, Oleksandr Holovachov1  \n \n1 Department of Zoology, Swedish Museum of Natural History, Box 50007, SE-104 05 \nStockholm, Sweden \n2 Plant Sciences Unit, Flanders Research Institute for Agriculture, Fisheries and Food (ILVO), \nBurg. Van Gansberghelaan 96, 9820 Merelbeke, Belgium \n \nCorresponding author: Oleksandr Holovachov (oleksandr.holovachov@nrm.se) \nAbstract \nNematode taxa of the Store Mosse National Park in the south of Sweden were surveyed using \nDNA metabarcoding. Samples were collected from a range of media across all the five vegetation \ntypes the park spans. A total of 50 samples consisting of soi l, litter, lichens, sphagnum, roots, \nwood, moss, fungus and anthill materials were analysed. Nematodes were characterised using a \n~350 bp region of their 18S ribosomal RNA gene that include V7 and V8 variable domains. The \nanalysis identified 47 families, 7 6 genera (21 new to Swedish fauna) and 60 species (31 new to \nSwedish fauna). Some nematodes showed a strong association with certain medium types, \nespecially at the species level. The results showed a strong justification for our strategy of sampling \ndifferent medium types. Soil and litter communities, which were the most diverse, showed high \nlevels of stability with good balance of all the various trophic and coloniser-persister groups. \nKeywords \nlitter, molecular marker, national park, Nematoda, soil, vegetation \n  \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nIntroduction \nNematodes represent a highly species -rich group that occur across a wide range of habitats with \nastonishing abundance (Holterman et al., 2006; van den Hoogen et al. 2019). There exists a great \namount of morphological, genomic and func tional diversity among nematodes, allowing them to \nplay diverse roles in the ecosystem. The variety of ways in which they respond to conditions of \ntheir environment is well documented and constitutes an important area of nematological research \n(Bongers 199 0). In fact, local nematode communities often reflect the prevailing physical, \nclimatic, biogeographical and chemical conditions of their environment (Cobb 1915; Yeates 1984; \nNeher 2001). Because of this, nematodes have been used as effective biological en tities for \nassessing the conditions of their environment. This is typically accomplished through \nmorphological identification of individual nematodes within the community to the levels of family \nor genus which are subsequently assigned to functional groups . However, morphological \nidentification of bulk nematode samples is time-consuming and requires expertise, the availability \nof which has been on a decline for some years now (Coomans 2002). On a small scale of a few \nsamples, this can easily be carried out by a skilled taxonomist or with the help of a good \nidentification key, by a keen observer of morphological characters to at least the family level. For \nsurvey studies involving tens to hundreds of samples, however, analysing and characterising \nnematodes in each sample quickly becomes an almost insurmountable undertaking. Molecular \nidentification methods provide faster and more accurate alternatives that require very little to no \ntaxonomic expertise and can be easily automated (Blok 2005; Ahmed et al. 2016).  Most popular \namong these was DNA barcoding, which involves the use of a targeted DNA region for \ndiscriminating species (Floyd et al. 2002; Blaxter 2003; Hebert et al. 2003; Blaxter et al. 2005). \nEarly studies on DNA barcoding using the Sanger sequencing approach were constrained in their \nthroughput – the number of species they can identify within a given time or sequence run is limited \n(Creer et al. 2010). For this reason, they have very limited practical application for assessing \nnematode communities. This limitation was finally overcome with the advent of next -generation \nsequencing (NGS) technology (Creer et al. 2010; Taberlet et al. 2012). \n  \nOne exciting concept that emerged because of NGS technology is DNA metabarcoding. \nMetabarcoding involves the use o f a standardised, typically short genomic region to characterise \norganisms from bulk samples. This approach has revolutionised the way we tackle questions \nrelated to biodiversity assessment (Guardiola et al. 2015). With its promise of providing a speedy \nmethod of assessing community structure, metabarcoding quickly saw wide adoption across many \nfields of meiofaunal studies (Porazinska et al. 2010a; Bik et al. 2012). Within nematology, most \nof the earlier studies on metabarcoding sought to evaluate the perfo rmance of different genomic \nregions and to establish robust pipelines for analysing bulk nematode samples (Porazinska et al. \n2009; Creer et al. 2010). Porazinska et al. (2009) assessed the suitability of metabarcoding for \nnematode community analysis and es tablished a benchmark for future nematode metabarcoding \nstudies. The authors used “mock” nematode communities to test the ability of the markers \n(NF1/18Sr2b of the SSU rDNA, D3Af/D3Br of the LSU rDNA) to recover the sampled taxa in \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nthese mock communities. The SSU rDNA-based primer they described was well regarded for its \nwide taxonomic coverage (Ahmed et al. 2019). While very useful in samples containing mostly \nthe targeted group, for samples with mixed taxa including non -targets, the broad taxonomic \ncoverage of these primers proved to be an issue. Because of this, there have recently been several \nattempts to develop alternative primers within the nuclear rDNA (Sapkota & Nicolaisen 2015; \nWaeyenberge et al. 2019; Kenmotsu et al. 2020; Sidker et al. 2020; Kawanobe et al. 2021). Similar \nbenchmarking studies have been carried out on nematodes in freshwater and marine habitats \n(Holovachov et al. 2017; Macheriotou et al. 2019; Schenk et al. 2020). \n  \nAn aspect of metabarcoding that has been the subject of several mock community studies concerns \nthe utility of sequence abundance information for inferring species abundance (Amend et al. 2010; \nPorazinska et al. 2010a; Elbrecht & Leese 2015; Thomas et al. 2016; Lamb et al. 2019). Most of \nthese studies have reported varyi ng degrees of divergence between taxa abundance and read \nfrequencies. In other words, the most abundant taxa do not often give the most reads and vice \nversa. Comparing biomass, instead of abundance, to sequence read frequency, however, has been \nshown to re sult in a better correlation for multiple marker regions (Schenk et al. 2020). Several \nfactors have been implicated as leading causes of this bias. These include primer mismatch, \nbiomass or size difference between individuals, DNA extraction bias where cuticles of some taxa \nhamper efficient tissue lysing, PCR bias, and for repetitive regions, copy number differences \n(Amend et al. 2010; Bik et al. 2013; Deagle et al. 2013). Quality filtering of reads during the \nbioinformatic analysis step can also contribute  to the discordance between sequence read \nfrequencies and taxa abundance. Attempts at mitigating this issue have taken different forms. In \ndiatoms, for example, a correction factor based on cell biovolume has been applied to minimise \nthe magnitude of the deviation between sequence reads abundance and taxa abundance (Vasselon \net al. 2018). Methods that involve gene enrichment and no PCR amplification have also been \nsuggested to eliminate all PCR-associated bias (Zhou et al. 2013). Using this approach in a st udy \non freshwater macroinvertebrates, Dowle et al. (2016) were able to obtain a strong correlation \nbetween biomass and read abundance. In addition to being more accurate at quantifying taxa, this \napproach may be the way to recover taxa which cannot be amplified by the current primer set. The \nconstraint here, however, is the increased cost and workload that may come with this method \n(Krehenwinkel et al. 2017). The use of presence -absence data as a safe and reliable substitute for \nabundance data is common. Here, since taxa occurrence and not read counts of taxa are used, read \nproliferation due to PCR bias does not have any effect. Of course, the biggest limitation to this \napproach is it ends up inflating the influence of very rare taxa and ignoring the fact th at some of \nthe differences in sequence read numbers have real biological bases (Deagle et al. 2019). \nMoreover, most nematode community indices can only be applied with data on abundance and \nhence have limited use for presence -absence data. Despite the subs tantial efforts in resolving the \nissue of abundance, there is currently no practical approach to predicting abundance from the read \nabundance. But does this suggest that read abundance data have no use at all? According to Deagle \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\net al. (2019), in spite of their sensitivity to recovery bias, relative read frequencies provided a more \naccurate representation of the population level diversity, compared to presence/absence data. \n  \nAs in other fields, metabarcoding has seen various applications in studying nematode biodiversity. \nPorazinska et al. (2010b) were among the first to apply metabarcoding to study the diversity of \nnematodes. Using metabarcoding, they examined the diversity of nematodes within different \nhabitats, soil, litter and canopy of Costa Rica. They concluded that nematode diversity was higher \nin tropical regions than in temperate regions, contrasting a previously held notion that suggested \nthe contrary. As a follow -up to this study, the main authors published another paper where they \ndemonstrated that nematode species richness in the tropical rainforest was three times more than \nit is in temperate rainforest, thus further supporting their earlier finding (Porazinska et al. 2012). \nIn their contribution to understanding the latitudinal differences in the diversity of nematodes, \nKerfahi et al. (2016) also used a metagenetic approach to examine nematodes from the tropical \nrainforest of Malaysia and the arctic tundra of Svalbard. They observed no difference between the \ntwo ecosystems. The fact that only s oil samples were used in their analysis may explain this \nobserved difference with Porazinska et al. (2010b; 2012). It is often rare in applied studies to have \nboth morphological and metabarcoding used complementarily, but Treonis et al. (2018) combined \nthe two approaches in a study on nematode communities under different cropping systems. Their \nresults indicated that metabarcoding provided better resolution beyond the family level, recovering \nfamilies that were not detected with morphological analysis. Metabarcoding also resulted in under- \nor overestimating the prevalence of some nematode families. \n \nA strong consensus exists across most studies regarding the ability of metabarcoding to recover \nmore taxa and reveal deeper taxonomic resolution when compared to the traditional morphological \napproach. We aim to leverage this by using metabarcoding to reveal the extent of nematode \ndiversity within the Store Mosse National Park in the south of Sweden. To our knowledge, this is \nthe first study into the nematode dive rsity of this park. The park spans five different vegetation \ntypes and at least eight different soil types. We sampled across all five vegetation types, with some \nmore heavily sampled than others because they were easily accessible. To better capture the \ndiversity of nematodes, we collected not just mineral soil, but other media such as litter, lichens, \nsphagnum, roots, decomposed wood, moss, fungus and samples from below anthills. Sampling \nacross vegetation types or medium types was not carried out evenly.  Therefore, the number of \nsamples varied across these two variables. Our goal with this study was also to use the nematode \ncommunity structure to infer the status of the different areas of the park, in terms of how pristine \nor perturbed they are. Given its  protected nature, we hypothesise that conditions will generally \nincline toward the former. \n \nMaterial and methods \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nStudy site \nAll samples were collected within the Store Mosse  National Park, located in the traditional \nprovince of Småland or present-day county of Jönköping situated in the southern part of Sweden \n(Fig. 1). Store Mosse is one of the largest bog complexes in southern Sweden (Martínez Cortizas \net al. 2021), covering approximately 7,682 hectares, most of which is wetlands. The bog consists \nmostly of high swamps with a few areas of open swamps (Naturvårdsverket 2015). Standing \nbetween 160–170 m above sea level, the area records an annual average of +6ºC of temperature \nand 766 mm of precipitation (Kylander et al. 2013). Store Mosse became a national park in 1982 \nfollowing a long campaign initiated by Prof. Edvard Wibeck for the area to be protected. The peat \nlayers covering its high swamps are believed to have accumulate d over a period of nearly 10,000 \nyears (Ryberg et al. 2022).  \nSample collection and processing \nSampling was carried out in 2021 over two days, the 13th and 14th of October. Samples were \ncollected from 50 spots across all vegetation types in the Store Mosse  National Park (Figs 1, 2, \nTable 1). Samples consisted of nine different types of media (Table 2). These diverse types of \nmedia were chosen to capture even those taxa found only in certain specific environments, and not \nin the commonly sampled mineral soil . At the same time, all samples were collected not too far \nfrom and along the roads and trails, in order to minimise our impact on undisturbed habitats. \nWhenever possible, samples were collected using a corer with inner diameter of 16 mm, collecting \n100 ml for each sample. Samples were then stored at 6ºC until extraction. Dense samples were \nmanually disintegrated prior to extraction in order to facilitate nematode isolation. Nematodes \nwere extracted from 100 ml of media using the Whitehead tray method (Whitehead and Hemming \n1965). The set -up was taken down after 48 hours and the extracts were collected in water \nsuspensions. Through a series of centrifugation, the suspensions were reduced to volumes of about \n50 µl inside microcentrifuge tubes. Each sample at this stage contained, in about 50 µl volume of \nsuspension, total assemblage of individuals of different taxa obtained from the extraction. The \nreduced volume of suspension was used to limit the chances of non -metazoan eukaryotes \nbecoming dominant in the samples. \n \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\n \n  \nFigure 1. Map of Store Mosse National Park (Sweden) showing all sampling points and the \ndifferent vegetation covers. \n  \n  \n \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nFigure 2. Images showing the different habitats where samples were collected. A. Lingonberries \nvegetation under a Pine forest. B. Pine forest with lichen and moss ground cover. C. Pine forest \nwith lingonberries, sphagnum and other bushes as ground cover. D. Fir and birch forest with litter \ncovering the ground surface. E. Open area within a pine forest covered with sedge ground cover. \nF. The bank of an artificial channel within a coniferous forest. G. Grassland with lower sphagnum \ncover. H. Granite outcrops.  \n  \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nTable 1. Sampling data. \n \nSample GPSLatitude GPSLongitude Vegetation Lower vegetation Medium \nSM01 57°14'50.66\"N 13°52'15.58\"E Coniferous forest on dry land Blueberries litter \nSM02 57°14'50.30\"N 13°52'16.46\"E Coniferous forest on dry land Blueberries litter \nSM03 57°14'52.84\"N 13°52'17.03\"E Coniferous forest on dry land Lingonberries litter \nSM04 57°14'54.68\"N 13°52'19.58\"E Coniferous forest on dry land Lingonberries litter \nSM05 57°14'59.48\"N 13°52'23.02\"E Coniferous forest on dry land Lingonberries soil \nSM06 57°15'5.83\"N 13°52'30.97\"E Coniferous forest on dry land Lingonberries soil \nSM07 57°15'6.01\"N 13°52'29.57\"E Coniferous forest on dry land N/A lichens \nSM08 57°15'8.77\"N 13°52'29.54\"E Coniferous forest on high swamp Blueberries soil \nSM09 57°15'34.06\"N 13°52'36.29\"E High swamp Grass sphagnum \nSM10 57°15'35.27\"N 13°52'47.30\"E Coniferous forest on dry land None litter \nSM11 57°15'41.39\"N 13°52'37.03\"E Coniferous forest on high swamp Blackberries litter \nSM12 57°15'43.47\"N 13°52'27.28\"E Coniferous forest on dry land Lingonberries litter \nSM13 57°15'40.00\"N 13°52'22.00\"E Coniferous forest on dry land Juncus/Carex roots \nSM14 57°15'40.57\"N 13°52'23.06\"E Coniferous forest on dry land N/A lichens \nSM15 57°15'39.04\"N 13°52'22.61\"E Coniferous forest on high swamp N/A wood \nSM16 57°15'47.83\"N 13°52'22.07\"E Coniferous forest on dry land none litter \nSM17 57°15'57.15\"N 13°52'33.17\"E Coniferous forest on dry land Blackberries fungus \nSM18 57°16'0.75\"N 13°52'32.96\"E Coniferous forest on high swamp N/A wood \nSM19 57°16'12.72\"N 13°52'44.62\"E Coniferous forest on high swamp Lingonberries litter \nSM20 57°15'48.75\"N 13°52'38.81\"E Coniferous forest on dry land Grass soil \nSM21 57°16'30.28\"N 13°53'57.63\"E Coniferous forest on dry land Grass sphagnum \nSM22 57°16'25.93\"N 13°54'21.23\"E Coniferous forest on high swamp Calluna litter \nSM23 57°16'25.30\"N 13°54'23.15\"E Coniferous forest on dry land N/A moss \nSM24 57°16'41.43\"N 13°54'28.43\"E Coniferous forest on dry land N/A litter \nSM25 57°17'10.01\"N 13°55'15.29\"E High swamp Crow berry litter \nSM26 57°17'12.49\"N 13°55'26.57\"E High swamp Crow berry litter \nSM27 57°17'16.91\"N 13°55'51.50\"E Coniferous forest on dry land Lycopodium soil \nSM28 57°19'35.22\"N 13°59'17.21\"E Coniferous forest on high swamp Crow berry litter \nSM29 57°19'15.11\"N 13°58'46.81\"E High swamp Carex litter \nSM30 57°19'15.62\"N 13°58'48.33\"E High swamp N/A wood \nSM31 57°19'0.41\"N 13°58'25.33\"E Deciduous forest Grass soil \nSM32 57°19'1.31\"N 13°57'58.73\"E Coniferous forest on high swamp Cranberries litter \nSM33 57°19'3.24\"N 13°57'44.68\"E Coniferous forest on high swamp Equisetum soil \nSM34 57°19'3.68\"N 13°57'42.74\"E Coniferous forest on high swamp N/A anthill \nSM35 57°19'7.82\"N 13°57'16.96\"E Coniferous forest on high swamp Blackberries moss \nSM36 57°19'11.44\"N 13°57'15.84\"E Coniferous forest on high swamp Blackberries wood \nSM37 57°19'9.45\"N 13°57'9.97\"E Coniferous forest on dry land N/A anthill \nSM38 57°19'8.03\"N 13°56'45.18\"E Coniferous forest on dry land Blackberries litter \nSM39 57°19'7.09\"N 13°56'29.78\"E Coniferous forest on dry land Blackberries fungus \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nSM40 57°18'58.91\"N 13°56'5.45\"E Deciduous forest Grass soil \nSM41 57°18'56.86\"N 13°56'0.07\"E Deciduous forest N/A lichens \nSM42 57°18'41.73\"N 13°56'10.24\"E Open swamp Semiaquatic plants soil \nSM43 57°18'29.40\"N 13°56'21.44\"E Coniferous forest on dry land N/A fungus \nSM44 57°18'18.85\"N 13°56'0.85\"E High swamp N/A moss \nSM45 57°18'16.89\"N 13°55'59.73\"E High swamp N/A wood \nSM46 57°18'20.36\"N 13°55'51.59\"E Coniferous forest on high swamp N/A soil \nSM47 57°18'20.98\"N 13°55'53.82\"E Coniferous forest on high swamp N/A soil \nSM48 57°18'11.92\"N 13°55'44.20\"E Coniferous forest on dry land Blackberries moss \nSM49 57°17'52.02\"N 13°55'53.66\"\nE \nHigh swamp N/A wood \nSM50 57°17'30.45\"N 13°56'33.55\"E High swamp Grass litter \n \nTable 2. Summary of vegetation covers of sampled sites and the types of media collected. \n \nMedium \nVegetation \nConiferous forest on dry land Coniferous forest on high \nswamp \nDeciduous \nforest \nHigh swamp Open \nswamp \nLitter SM1, SM2, SM3, SM4, SM10, \nSM12, SM16, SM24, SM38 \nSM11, SM19, SM22, SM28, \nSM32 \n SM25, SM26, SM29, \nSM50 \n \nSoil SM5, SM6, SM20, SM27 SM8, SM33, SM46, SM47 SM31, SM40  SM42 \nLichens SM7, SM14  SM41   \nSphagnum SM21   SM9  \nRoots SM13     \nDecomposed \nwood \n SM15, SM18, SM36  SM30, SM45, SM49  \nMoss SM23, SM48 SM35  SM44  \nFungus SM17, SM39, SM43     \nAnthill SM37 SM34    \n \nDNA Extraction, PCR and NGS \nGenomic DNA extraction was performed on each sample using the Qiagen QiAmp  DNA Micro \nkit. Briefly, 130 µl of ATL buffer was added to each sample (50 µl), followed by the 20 µl of \nproteinase K. The mixture was then vortexed and incubated overnight in an incubating microplate \nshaker at 56 ºC and 300 rpm. Pure DNA was isolated from  the lysed samples following the \nmanufacturer’s instructions for genomic DNA extraction for blood and tissue samples using the \nQiagen QiAmp DNA Micro kit. The PCR primers used were the NF1 (5 \n‘GGTGGTGCATGGCCGTTCTTAGTT 3’, matching the 5’ end of the 38th helix) and 18Sr2b (5’ \nTACAAAGGGCAGGGACGTAAT 3’, matching the 3’ end of the 32nd helix) (Porazinska et al. \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\n2009), with Illumina adapter sequences ligated at their 5’ ends. Amplification was performed in a \n25 µl reaction mixture using Illustra Hot Start Mix RT G0.2 ml reaction kit (GE Healthcare Life \nSciences, Sweden). The reaction mixture consisted of 1 µl (0.4 µM) of each primer, 2 µl of \ntemplate DNA and 21 µl of nuclease -free water. The cycle conditions set were as previously \ndescribed in Ahmed et al. (2019).  Following the initial PCR reaction, the amplicons were all \npurified using Agencourt AMPure XP (Beckman Coulter, California, United States). The purified \nproducts were sent to Macrogen Europe B.V. (Amsterdam, the Netherlands) for subsequent library \npreparation and NGS. In brief, index PCR was performed where unique indices were attached to \namplicons of each sample. Library size distribution was checked by running on Agilent \nTechnologies 2100 Bioanalyzer using a DNA 1000 chip. Quantification of the library w as \nperformed using qPCR according to the Illumina qPCR quantification protocol guide. The samples \nwere then multiplexed, before Illumina MiSeq 2x300 bp sequencing. Raw data is available at \nNCBI Sequence Read Archive under the BioProject ID PRJNA923582. \nBioinformatics \nAnalysis of the raw NGS data was performed using a 64-bit USEARCH v11.0.667 (Edgar 2010), \nlicensed to the senior author (academic non -profit licence). Raw reads were merged using the  -\nfastq_mergepairs command (options: the minimum length of ove rlap between the forward and \nreverse reads was set to 150 bp, the maximum number of mismatches within the overlapping region \nset to 10, the minimum percentage identity at the overlapping region set to 80). Following this, the \nmerged reads were filtered using the usearch command -fastq_filter (options: maximum expected \nerror per sequence was set to 1, minimum length of the sequences set to 250). Using the  -\nfastx_uniques command, the remaining reads were reduced to unique sequences in order to speed \nup the clustering step. The output file consisted of unique sequences, each with its size appended \nto its description line. This was followed by clustering using the UNOISE algorithm, as \nimplemented in the  -unoise3 command, to obtain amplicon sequence variants (ASV s), also \nreferred to as zero-radius operational taxonomic units (ZOTUs) by Edgar (2016). The use of ASVs \nhas been shown to recover a higher number of correct biological sequences than the UPARSE -\nOTU algorithm. All parameters for the clustering command were  set to their default values. This \nincluded a -minsize value of 8 which ensured that only ZOTUs with an abundance of 8 or higher \nwere retained. Low -minsize values tend to introduce more errors in the predicted low-abundance \nbiological sequences (see https://www.drive5.com/usearch/manual/cmd_unoise3.html). The \nusearch command -otutab was then used to create a ZOTU table, a table with ZOTUs shown as \nrecords, and their read frequencies in each sample shown in separate fields. Using the  -sintax \ncommand and 18S rDNA sequences from the PR2 database version 4.14.0 (Guillou et al. 2013) as \nreference, the ZOTUs were assigned taxonomy. A pattern matching script was used to extract \nZOTUs matching Nematoda from the sintax taxonomy assignment. These ZOTUs were then \nsearched for and extracted from the ZOTU sequences and ZOTU table. The extracted nematode \nZOTUs were then assigned a taxonomy, this time using a custom high -quality nematodes-only \ncurated database with more accurate taxonomies based on the classification of De Ley and Blaxter \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\n(2004). Initially, the high posterior probability score cut -off set for the sintax taxonomy resulted \nin ZOTUs associated with Dorylaimidae and Qudsianematidae failing to be assigned to the correct \nfamilies. Blast search results were theref ore used to complement the sintax taxonomy. A \nphylogenetic tree of the nematodes was generated for mafft -aligned (Katoh and Standley 2013) \nZOTUs using FastTree (Price et al. 2010). All parameters were left at their default settings for \nmafft alignment, and for FastTree analysis, the ‘gtr’ model was used. The ZOTU table, taxonomy \nfile, phylogenetic tree and sample metadata file (data on the vegetation type, medium, soil type etc \nfor each sample) were exported into R for further statistical analyses. \nStatistical analyses \nComputation of community indices, CP (coloniser -persister) groupings and feeding group \ndesignations were performed using NINJA (Sieriebriennikov et al. 2014; \nhttps://shiny.wur.nl/ninja/), an online tool for nematode faunal analyses. Taxa not recognized by \nNINJA were replaced by their closest relative acceptable to NINJA or where possible replaced by \na higher -ranking taxon (e.g. Basilaphelenchus replaced by Aphelenchoididae). Abund ance \ninformation was based on sequence read counts for each taxon in a sample. For comparisons, \nsamples were categorised in terms of the different vegetation types and the different media from \nwhich the nematodes were collected. All analyses were carried out using R version 4.0.5 (R Core \nTeam 2021) inside RStudio (RStudio Team 2022). Alpha diversity within the vegetation types and \nthe medium types were computed based on Chao1 diversity index using the phyloseq package \nMcMurdie and Holmes (2013). \nResults \nGeneral statistics  \nA total of 7,040,489 paired reads were generated. On average 85% of the paired reads were \nsuccessfully merged per sample. Following filtering, 5,943,062 reads were left. Clustering using \nUNOISE and setting the -minsize parameter to 8 resul ted in a total of 2,569 ZOTUs. The sintax \nalgorithm assigned 31.8% (899 in total) of the ZOTUs to Nematoda, 18.7% were unassigned, and \nthe remainder were assigned to other eukaryotic lineages (Fig 3). By design, each taxonomic rank \nassignment in the sintax output is accompanied by a posterior probability value, which indicates \nthe statistical support for that particular assignment. For this analysis, only assignments with \nposterior probability scores of >=0.8 were considered. The primer pair used for the DN A \namplification has been shown to demonstrate a wide taxonomic coverage –capable of amplifying \neven Archaean DNA. However, because nematodes were first isolated from the medium most of \nthe non-target DNA were excluded, that otherwise would have dominated th e samples had DNA \nextraction been performed directly on the sample medium. Furthermore, getting rid of most of the \nsuspension by concentrating the samples to about 50 µl ensured that the incidence of fungal DNA \nwas significantly suppressed. The nematodes extraction method used also recovered a noticeable \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nnumber of tardigrades, platyhelminths, rotifers and arthropods (Fig. 3), which, however, were not \nanalysed in details. \n  \n  \n  \n \nFigure 3. Relative abundance of ZOTUs associated with major groups of eukaryotes in all 50 \nsamples combined, including ZOTUs that were not assigned to any group of eukaryotes. \nNematode diversity  \nThe analysis recovered 46 nematode families in total, with Tylenchidae represented by the greatest \nnumber of ZOTUs (Fig. 4). At th e genus level, the percentage of nematode ZOTUs that were \nassigned taxonomy with enough support averaged about 40% and in some samples was as high as \n78%. Across the 47 families recovered, there were 76 genera identified in the analysis (Table 3). \nThe most  represented families were Aphelenchoididae (7 genera), Rhabditidae (8 genera) and \nTylenchidae (11 genera). Most of the families had more than one genus present. Identification to \nthe species level was accomplished for 46 of the genera. Sixty nematode spec ies were identified \nin total representing all five trophic groups. Aphelenchoides and Malenchus were the most diverse \ngenera with four species associated with both. Among these 60 identified species, 31 are new to \nthe fauna of Sweden (Dyntaxa): Acrobeloides varius, Aglenchus agricola, Aphelenchoides \nblastophthorus, Aphelenchoides heidelbergi, Baldwinema ardabilense, Cephalenchus \nhexalineatus, Choriorhabditis cristata, Crassolabium circuliferum, Diploscapter coronatus, \nDitylenchus adasi , Ecphyadophora tenuis sima, Filenchus facultativus, Helicotylenchus \npseudorobustus, Hexatylus viviparus, Irantylenchus vicinus, Laimaphelenchus penardi, \nMalenchus bryanti, Malenchus neosulcus, Malenchus pressulus, Miculenchus muscus, Oscheius \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\ndolichura, Paravulvus hartingii, Pa urodontella gilanica, Potensaphelenchus stammeri, \nTylencholaimus teres, Tylencholaimus zhongshanensis, Tylenchorhynchus parvulus, Tylenchus \narcuatus, Tylenchus naranensis, Tylolaimophorus typicus, Veleshkinema iranicum. Similarly, the \nfollowing 21 genera h ave not been reported in Sweden until now (Dyntaxa):  Baldwinema, \nBasilaphelenchus, Choriorhabditis, Crassolabium, Diploscapter, Discotylenchus, \nEktaphelenchoides, Heterorhabditis, Hexatylus, Irantylenchus, Laimaphelenchus, Miculenchus, \nNeodolichorhynchus, Oscheius, Paravulvus, Paurodontella, Poikilolaimus, Potensaphelenchus, \nRhabditophanes, Schistonchus, Veleshkinema. \n \nFigure 4. Proportions of total ZOTUs assigned to various nematode families including those \nunassigned at the family level across all samples. \n \nTable 3. List of genera and species of nematodes identified across all samples. The families for \nwhich genus assignment could not be achieved are not represented in this table. Number of ZOTUs \nidentified for each taxon in parenthesis. \nFamily Genus Species \nActinolaimidae (2) Paractinolaimus (1)   \nAlaimidae (12) Alaimus (4) Alaimus parvus (1) \nAlloionematidae (1) Rhabditophanes (1)   \nAngiostomatidae (1) Angiostoma (1) Angiostoma margaretae (1) \nAnguinidae (25) Anguina (4)   \nDitylenchus (14) Ditylenchus adasi (1), D. destructor (1) \nAphanolaimidae (9) Aphanolaimus (4) Aphanolaimus aquaticus (2) \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nAphelenchoididae (115) Aphelenchoides (68) Aphelenchoides blastophthorus (1), A. heidelbergi (5), A. \nritzemabosi (3), A. saprophilus (1) \nBasilaphelenchus (15)   \nBursaphelenchus (1)   \nEktaphelenchoides (2)   \nLaimaphelenchus (16) Laimaphelenchus penardi (6) \nPotensaphelenchus (3) Potensaphelenchus stammeri (1) \nSchistonchus (1)   \nAporcelaimidae (11) Aporcelaimellus (9) Aporcelaimellus obtusicaudatus (7) \nBunonematidae (5) Bunonema (5) Bunonema reticulatum (1), B. richtersi (1) \nCephalobidae (6) Acrobeloides (4) Acrobeloides varius (1) \nChronogastridae (6) Chronogaster (6)   \nDesmodoridae (86) Prodesmodora (86)   \nDiphtherophodridae (8) Diphtherophora (1)   \nTylolaimophorus (5) Tylolaimophorus typicus (5) \nDiplogastridae (5) Pristionchus (3)   \nDiplopeltidae (1) Cylindrolaimus (1)   \nDorylaimidae (5) Crassolabium (1) Crassolabium circuliferum (1) \nProdorylaimus (1)   \nEthmolaimidae (1) Ethmolaimus (1) Ethmolaimus pratensis (1) \nHoplolaimidae (2) Helicotylenchus (1) Helicotylenchus pseudorobustus (1) \nMetateratocephalidae (21) Metateratocephalus (16) Metateratocephalus crassidens (7) \nEuteratocephalus (2) Euteratocephalus palustris (1) \nMonhysteridae (61) Eumonhystera (59) Eumonhystera filiformis (3) \nGeomonhystera (2)   \nMononchidae (25) Clarkus (3) Clarkus papillatus (2) \nMononchus (10) Mononchus truncatus (7) \nPrionchulus (9) Prionchulus muscorum (9) \nMylonchulidae (1) Mylonchulus (1)   \nNeotylenchidae (4) Hexatylus (4) Hexatylus viviparus (3) \nNordiidae (11) Enchodelus (5)   \nPungentus (1)   \nNygolaimidae (6) Paravulvus (2) Paravulvus hartingii (2) \nPanagrolaimidae (13) Baldwinema (1) Baldwinema ardabilense (1) \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\n Panagrolaimus (7)   \nPlectidae (34) Plectus (25) Plectus minimus (1), P. tenuis (6) \nTylocephalus (2) Tylocephalus auriculatus (1) \nPratylenchidae (1) Pratylenchus (1) Pratylenchus crenatus (1) \nPrismatolaimidae (8) Prismatolaimus (8) Prismatolaimus dolichurus (3) \nRhabditidae (20) Choriorhabditis (1) Choriorhabditis cristata (1) \nDiploscapter (1) Diploscapter coronatus (1) \nHeterorhabditis (1)   \nOscheius (1) Oscheius dolichura (1) \nPellioditis (1)   \nPoikilolaimus (2)   \nProtorhabditis (1)   \nRhabditis (10)   \nRhabdolaimidae (3) Rhabdolaimus (3)   \nSphaerulariidae (5) Paurodontella (2) Paurodontella gilanica (2) \nVeleshkinema (2) Veleshkinema iranicum (2) \nSteinernematidae (1) Steinernema (1) Steinernema kraussei (1) \nTelotylenchidae (1) Neodolichorhynchus (1)   \nTylenchorhynchus (2) Tylenchorhynchus parvulus (2) \nTeratocephalidae (7) Teratocephalus (7) Teratocephalus deconincki (2) \nTrichodoridae (1) Paratrichodorus (1) Paratrichodorus pachydermus (1) \nTripylidae (3) Tripyla (2)   \nTylenchidae (151) Aglenchus (1) Aglenchus agricola (1) \nBasiria (1)   \nCephalenchus (3) Cephalenchus hexalineatus (2) \nCoslenchus (1) Coslenchus costatus (1) \nDiscotylenchus (1)   \nEcphyadophora (7) Ecphyadophora tenuissima (2) \nFilenchus (16) Filenchus facultativus (7), F. misellus (1) \nIrantylenchus (4) Irantylenchus vicinus (2) \nMalenchus (68) Malenchus acarayensis (4), M. bryanti (3), M. neosulcus (17), M. \npressulus (6) \nMiculenchus (24) Miculenchus muscus (5) \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nTylenchus (9) Tylenchus arcuatus (2), T. naranensis (1) \nTylencholaimidae (8) Tylencholaimus (6) Tylencholaimus mirabilis (3), T. teres (1), T. zhongshanensis (1) \nTylenchulidae (2) Paratylenchus (2)   \nXyalidae (10) Theristus (9) Theristus agilis (8) \nNematode communities across medium types and vegetations \nSoil and litter samples had high ZOTU richness. However, this is most likely because there were \nmore samples collected for these medium types. The two medium types showed comparable \nrichness (Fig. 5a). Moss samples were the closest to these two in terms of richness. Fungus, lichens, \nand decomposing wood samples, on the other hand, had low alpha diversity scores. The two most \nheavily sampled vegetations, coniferous forest on dry land and coniferous forest on high swamp \nshowed a wide range of diversity across sampled spots. Between them, there was a significant \ndifference in their alpha diversity (Fig. 5b). \n  \n \nFigure 5. Chao1 measures of α -diversity of nematodes in different samples. a) Medium types. \nStatistical significance of the difference between alpha diversity for litter and soil samples was \ntested using the Wilcoxon test. ns = not significant. b) Vegetation types. Statistical significance of \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nthe difference between alpha diversity for Coniferous Forest on dry land and Coniferous Forest on \nhigh swamp samples was tested using the Wilcoxon test. * = significant. \n  \n \nFigure 6. Read distribution among nematode families. Each bar corresponds to a sample. Samples \nare aggregated into various medium types (a) and vegetation types (b). Not all taxa were resolved \nto the species level. These are represented at the order or class rank.  \n \nZOTUs associated with Qudsianematidae were dominant in most of the samples, especially the \nlitter and soil samples (Fig. 6a). Rhabditidae and Plectidae dominated the two samples collected \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nfrom the anthills. Fungus samples produced very high number of reads  associated with \nDiplogastridae, Alloionematidae, Aphelenchoididae, unidentified Rhabditida and Plectidae. Aside \nfrom being the most dominant family in litter samples, Qudsianematidae was also the most \nprevalent, occurring in each of the eighteen litter sa mples (Fig. 6a). Sequence reads associated \nwith Plectidae, Mononchidae, Qudsianematidae and an unidentified Dorylaimida dominated the \nmoss samples. There were also a few Anguinidae and Metateratocephalidae reads present. In the \nsingle root sample collected, Tylenchidae, Prismatolaimidae, Mononchidae, Metateratocephalidae \nand Chronogastridae were dominant. Qudsianematidae was the most prevalent family in the soil \nsamples, absent only in two of eleven samples. It was also the dominant family in four of the so il \nsamples. Next in terms of prevalence was Tylenchidae, followed by Rhabditidae, then another \ndorylaimid family, Tylencholaimidae. The taxonomic compositions of the two sphagnum samples \nwere similar, except for Teratocephalidae and Xyalidae, which were fo und only once with \nsignificant abundance. As expected, the family Aphelenchoididae dominated the wood samples. \nOne of the wood samples had a substantial number of reads belonging to Anguinidae, Rhabditidae \nalong with an unidentified Rhabditida. Tylenchidae  was also found in three of the six wood \nsamples. Nematode distribution across samples within the same vegetation types did not show \nsame level of similarity observed in samples from the same medium types, indicating a lack of \ncorrelation between community structure and vegetation type (Fig. 6b). \n \nIn terms of prevalence, most taxa showed wide distribution across multiple medium types and \nvegetations (Figs 7, 8). Rhabditis (soil), Poikilolaimus (wood), some species of Basilaphelenchus \n(wood) and an unidentif ied Dorylaimida (soil) were confined to only one medium type. \nHeterorhabditis (coniferous forest on dry land), Criconematidae (coniferous forest on dry land), \nsome unidentified species of Dorylaimida (coniferous forest on high swamp) and unidentified \nTriplonchida (coniferous forest on dry land) were also associated with just one vegetation type. \n \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\n  \nFigure 7. Maximum likelihood tree of the 100 most dominant ZOTUs showing their prevalence \nand abundance across different samples and medium types. Leaf nodes are labelled with the \nassigned taxa (genus where possible) of the ZOTUs. Circles represent the samples, and the \ndiameters of the circles indicate the abundance of the taxon in samples. Circles of the same colour \nindicate samples from the same medium type. \n  \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\n  \n  \nFigure 8. Maximum likelihood tree of the 100 most dominant ZOTUs showing their prevalence \nand abundance across different samples and vegetation types. Leaf nodes are labelled with the \nassigned taxa (genus where possible) of the ZOTUs. Circles represent the samples, and the \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\ndiameters of the circles indicate the abundance of the taxon in samples. Circles of the same colour \nindicate samples from the same vegetation type. Con. = Coniferous, Dec. = Deciduous. \nNematode trophic groups and coloniser-persister groups \nAll five major trophic groups described by Yeates et al. (1993) were recovered in each of the \nmedium types. Their distribution, however, varied across the medium types (Fig. 9a). Fungivores \nwere most dominant in decomposing wood samples, but  very low in moss, sphagnum, and root \nsamples. Bacterivores occurred with significant abundance across all sample types, particularly in \nanthill, lichen, moss, and root samples. The occurrence of predators was very low in lichen and \nsphagnum samples. Omnivores had high abundances in the samples associated with litter, moss, \nsoil, and sphagnum. They were also well represented in the wood samples. The occurrence of \nherbivores appeared to be in concert with that of fungivores in the different medium types, except \nfor decomposing wood samples where almost half of all reads came from the latter. Bacterivore -\nlinked c-p groups 1 and 2 dominated the reads in fungus, lichen, and decomposing wood samples \n(Fig. 9b). Fungus samples were dominated by c -p 1 nematodes, sug gesting high level of \nenrichment. Soil samples showed a uniform distribution of all five c -p classes, a feature of a \npristine or relatively undisturbed community. Litter samples also showed a c -p class distribution \nsimilar to that of soil samples. \n  \n  \nFigure 9. Distribution of nematode trophic and coloniser-persister groups in different medium and \nvegetation types. \n \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nThere was also a uniform representation of all trophic groups in coniferous forest on dry land, and \nto some extent in high swamp vegetation ty pes (Fig. 9c). The single sample representing open \nswamp was dominated by bacterivorous nematodes, representing over 60% of the nematode \nassemblage. Predators and omnivores together represented about a third of the assemblage. \nFungivores and herbivores were very few. In coniferous forest on high swamp and deciduous forest \nsamples, all the trophic groups were well represented except predators which occurred with very \nlow abundance in both vegetation types. The single open swamp sample was dominated by c -p 1 \nnematodes (Fig. 9d). All c -p classes, except c -p 5 were well represented in the high swamp \nsamples. In the two coniferous forest types, c-p 4 nematodes were the dominant group. The good \nstructure indicators, c -p 4 and c -p 5 nematodes constituted a high pro portion of the assemblage \nassociated with deciduous forest vegetation (Fig. 9d). \nAnalysis of disturbance levels and food web in the communities \nUsing the c-p triangle to depict the stability/enrichment/stress conditions of the communities, most \nsamples appeared to be in good stable conditions (Figs 10a, 10b). The soil and litter samples, where \nmost of the diversity occurred, showed the highest stability while at the same time showing very \nlow levels of enrichment. Wood and lichen samples were generally depi cted as stressed, except \nfor a few samples of decomposing wood. The two anthill samples were in low stability states. \nBased on the interpretation by Ferris et al. (2001), most of the samples regardless of the medium \ntype were either in a matured or maturing food web state (Fig. 10b). Lichen was the only medium \nfor which all samples were in a degraded, depleted state, with high C:N ratio. The two sphagnum \nsamples fell within the matured and fertile category, with moderate C:N ratio. Soil and litter \nsamples were mostly concentrated within the high structure quadrants (maturing to matured food \nweb), although with varying degrees of enrichment. Fungus samples showed low maturity and \nappeared to be highly disturbed in some samples and enriched in others. Their decomposition was \ngenerally not fungal but bacterial driven. The only root sample collected depicted a food web that \nwas matured and fertile. The wood samples varied greatly in the states of the food webs they \ndepicted, while some appeared matured and fertil e/N-enriched, others were highly disturbed and \nmoderate to heavily enriched. This can be explained by the fact that wood decomposition is a \ncomplex multistage process that includes different organisms during different times, with wood at \nlate stages becomi ng similar to litter and soil. The two anthill samples were in quite opposite \nconditions, one in a maturing state while the other was in a degraded and nutrient-depleted state.  \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\n \nFigure 10. CP triangle and food web analysis of the different medium types. CP triangles (a) depict \nthe stability of the communities. Food web analysis plots (b) depict the maturity of the food webs \nwithin the communities. a) CP triangle showing samples categorised under different medium \ntypes. b) Food web analysis showing samples categorised under different medium types.  \n  \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nDiscussion \nNematode suspension after extraction from substrates such as soil and litter often contain other \nmetazoans. Among the most frequently encountered groups in such nematode extracts are \nterrestrial tardigrades and some micro -arthropods. It was therefore not unexpected that almost a \ntenth of the total ZOTUs were assigned to these two groups combined, especially considering the \nuniversal nature of the primers used. Our effort to minimise the amplification of other non-targets \nsuch as fungi by extracting nematodes first from the substrate and reducing the volume of the \nsuspension down to 50µl was to some extent successful. But despite that, fungal and unidentified \nZOTUs (most likely Archaean) together consti tuted almost 38% of the total ZOTUs. Primer \ncombinations exist that can address this through preferential amplification of nematode 18S rDNA \n(Sapkota & Nicolaisen 2015; Waeyenberge et al. 2019; Kawanobe et al. 2021), thus limiting the \namplification of non -nematode DNA. Kawanobe et al. (2021) demonstrated through in silico \nanalysis the high tendency of NF1/18Sr2b primers to amplify non -nematode eukaryotes. Their \nanalysis identified three primer combinations that showed better coverage and specificity to \nnematodes. Our goal for this study was not to exclude all other eukaryotes. Therefore, using any \nof the nematode-specific primers would have limited the detection of other metazoans recovered \nin this study. For most use cases however, these nematode -specific p rimers can be better \nalternatives to the widely used NF1/18Sr2b. \n  \nOur analysis recovered a massive diversity of nematodes, with a total of 47 nematode families \nidentified representing 10 different orders. We identified 76 nematode genera in total across all \nsamples, 46 of which were identified to species level. Even thou gh reads associated with family \nQudsianematidae were the most dominant, species or genus assignments of ZOTUs belonging to \nQudsianematidae were not supported according to the sintax assignment method. Similar to the \nRDP naïve Bayesian classifier, the sinta x attaches posterior probability scores to each rank \nclassification (Wang et al. 2007; Edgar 2016). Any rank classification receiving a support value \nbelow the set threshold (0.8/1 in this study) was considered not supported enough. For some taxa \na blast s earch against the Genbank reference database could resolve the assignment. For orders \nsuch as Dorylaimida and Rhabditida, although more so for the former, many ZOTUs could not be \nidentified further beyond the rank of an order, even with BLAST search. In th e case of \nQudsianematidae, the ZOTUs could only be assigned to the rank of family (Figs 7; 8). A possible \nexplanation for this is the conserved nature of the 18S rDNA region for delineating some species. \nThis is particularly widespread among species of mos t free -living Dorylaimida of which \nQudsianematidae is a member. With this group, it has been shown that even the full -length 18S \nrDNA sequence is extremely conserved (Holterman et al. 2006; 2008). \n \nLitter and soil were the two most sampled habitats. For th e majority of the samples, these two \nmedium types shared similar taxonomic composition (Figs 6a, 6b). However, in some samples \nthere were observable differences in the community structure. For example some soil samples \nappeared to be uniquely dominated by Rhabditidae which were not observed with such high \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nabundance in the litter samples. Other than these two medium types which were generally similar \nin their community structure, the rest of the medium types produced unique taxonomic \ndistributions (Fig. 6a). At the species level, some taxa showed exclusive association with certain \nmedium types (Figure 11). In general, it appears that had the sampling effort been limited only to \nmineral soil samples, the majority of the observed nematode families would still h ave been \nrecovered. However, at the species level, many taxa would have been missing, specifically those \nassociated with litter, decomposing wood, sphagnum, lichens and moss samples. Soil sampling \nalone does not adequately reflect the diversity of nematode s, and depending on the geographic \nlocation, the degree to which taxa are missed as a result of this sampling strategy can vary (Powers \net al. 2009; Porazinska et al. 2010b; 2012). Even though the collection of samples beyond the \nmineral soils has in fact been practised by authors in the past (Yeates 1972; Sohlenius and Boström \n2001), and that the importance of collecting litter and other materials above the mineral soil was \nunderstood then, some biodiversity studies still focus only on soil samples. The cu rrent study \nconcurs with previous studies in demonstrating the importance of sampling not just the mineral \nsoil, but other habitats as well (Powers et al. 2009; Porazinska et al. 2010b; 2012). This must be \ntaken into consideration when attempting to create a baseline metabarcoding datasets for different \nbiomes to be used as reference points in the future monitoring of ecosystem changes. Such baseline \nreference datasets for complex land use type (forest) can not be based on standard soil+litter \nsamples. As w e have shown above, a considerable percentage of diversity was found in media \ntypes other than mineral soil and litter. Thus, in order to establish comprehensive metabarcoding \nbaseline for a given biome, all possible microhabitats where the target taxon ma y occur, must be \nincluded in sequencing. \n \nIn spite of the inability of the barcode marker to identify the Qudsianematid ZOTUs beyond the \nfamily level, the 76 genera recovered is quite remarkable, especially considering the fact that the \ncurrent study was b ased on a one -time sampling. Comparing this study with others carried out \nwithin Sweden or regions with similar climatic conditions clearly shows a higher recovery of \nnematode taxa. For example, over the course of three sampling series spanning over a period of 25 \nyears, in which 156 soil samples from Scots pine forest in Sweden were analysed, Sohlenius and \nBoström (2001) identified 36 unique nematode taxa. Of these, 31 were identified to the genus \nlevel. The majority of the genera identified in that study were recovered here as well, with the \nexception of Geocenamus, Acrobeles, Cervidellus, Achromadora, Wilsonema, Eudorylaimus, \nMicrodorylaimus and Thonus. In contrast, over 50 of the genera identified in this study were \nmissing in their taxa list. Compared to the current study, Yeates (1972) also identified fewer unique \ngenera (41) from an 85-year old Danish beech forest studied over a period of 12 months. Another \nstudy on the metazoan microfauna of a mire in northern Sweden sampled over a period of four \nmonths also identified 24 taxa representing 17 unique genera (Sohlenius et al. 1997). One of the \nlikely reasons for this is that metabarcoding allows to identify immature individuals and eggs, \nwhich morphology-based approaches are unable to do. \n \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\n  \nFigure 11. Species network showing association between taxa and samples. Samples represented \nby hexagonal nodes; taxa represented by circular nodes. Sample -taxon association depicted by \narrows extending from the sample to the taxon. Medium types are represented by different colours. \nBoth the nodes representing a sample and edge extending from it are colour to depict the medium \ntype it belongs to. \n \nNematode distribution appeared to show no association with vegetation types. It appeared \nnematode community structure was influenced more by the medium type that was collected. This \nwas expected given that most nematode taxa will be more associated with certain habitats than \nothers regardless of what the local vegetation cover of the habitat is. Across the different \nvegetations, none of the community structures stood out as unique (Fig. 6b). Due to the strong \ninfluence the medium type has on the community, a better comparison of communities under the \ndifferent vegetation types would be one that is restricted to only one type of sample medium. This \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\ncould not be done for any of the medium types because none of the medium types was represented \nacross all five vegetation types. \n  \nIndices used in this study that describe the structure and maturity of the community are heavily \ndependent on abundance data. And since sequence read abundance does not always directly \ncorrelate with the abundance of taxa in a typical metabarcoding analysis, there is constraint in the \ninferences that can be made about the condition of the samples based on t hese indices \n(Waeyenberge et al. 2019). Nevertheless, the largely stable conditions depicted by our analysis are \nexpected especially given the pristine nature of the Store Mosse National Park. \nAcknowledgements \nThis research was in part supported by the gra nt from the Stiftelsen Anna och Gunnar Vidfelts \nfond för biologisk forskning (2020 -071-Vidfelts fond/SOJOH) “Taxonomic and functional \ndiversity of Nematode fauna of the Store Mosse National Park: a metabarcoding approach” for \nMA and OH and by the European H2020 program through the SoildiverAgro -project grant \nagreement 817819 for DS. Sampling in the Store Mosse National Park was performed within the \npermit # 521-5933-2021 issued by the Länsstyrelsen i Jönköpings län. USEARCH v11.0.667 64 \nbit for OS X was used under the academic non-profit licence.  \nReferences \nAhmed M, Sapp M, Prior T, Karssen G, Back MA (2016) Technological advancements and their \nimportance for nematode identification. SOIL 2: 257 –270. https://doi.org/10.5194/soil-2-\n257-2016  \nAhmed M, Back MA, Prior T, Karssen G, Lawson R, Adams I, Sapp M (2019) Metabarcoding of \nsoil nematodes: the importance of taxonomic coverage and availability of reference \nsequences in choosing suitable marker (s). Metabarcoding and Metagenomics 3: e36408. \nhttps://doi.org/10.3897/mbmg.3.36408  \nAmend AS, Seifert KA, Bruns TD (2010) Quantifying microbial communities with 454 \npyrosequencing: does read abundance count? Molecular Ecology 19: 5555 –5565. \nhttps://doi.org/10.1111/j.1365-294x.2010.04898.x  \nBik HM, Sung WAY, De Ley P, Baldwin JG, Sharma J, Rocha -Olivares A, Thomas WK (2012) \nMetagenetic community analysis of microbial eukaryotes illuminates biogeographic \npatterns in  deep-sea and shallow water sediments. Molecular Ecology 21: 1048 –1059. \nhttps://doi.org/10.1111/j.1365-294x.2011.05297.x  \nBik HM, Fournier D, Sung W, Bergeron RD, Thomas WK (2013) Intra -genomic variation in the \nribosomal repeats of nematodes. PLoS One 8: e78230. \nhttps://doi.org/10.1371/journal.pone.0078230  \nBlaxter M (2003) Counting angels with DNA. Nature 421. https://doi.org/10.1038/421122a  \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nBlaxter M, Mann J, Chapman T, Thomas F, Whitton C, Floyd R, Abebe E (2005) Defining \noperational taxonomic units using DNA ba rcode data. Philosophical Transaction of the \nRoyal Society Lond B 360: 1935-1943. https://doi.org/10.1098/rstb.2005.1725  \nBlok VC (2005) Achievements in and future prospects for molecular diagnostics of plant-parasitic \nnematodes. Canadian Journal of Plant Pathology 27: 176 –185. \nhttps://doi.org/10.1080/07060660509507214  \nBongers T (1990) The maturity index: an ecological measure of environmental disturbance based \non nematode species composition. Oecologia 83: 14 –19. \nhttps://doi.org/10.1007/BF00324627  \nCobb NA (1915) Nematodes and their relationships. In: USDA Yearbook of Agriculture, 1914.  \nUS Government Printing Office, 457–490. \nCoomans A (2002) Present status and future of nematode systematics. Nematology 4: 573 –582. \nhttps://doi.org/10.1163/15685410260438836  \nCreer S, Fonseca VG, Porazinska DL, Giblin-Davis RM, Sung W, Power DM, Packer M, Carvalho \nGR, Blaxter ML, Lambshead PJD, Thomas WK (2010) Ultrasequencing of the meiofaunal \nbiosphere: Practice, pitfalls and promises. Molecular Ecology 19: 4 –20. \nhttps://doi.org/10.1111/j.1365-294X.2009.04473.x  \nDe Ley P, Blaxter ML (2004) A new system for Nematoda: combining morphological characters \nwith molecular trees, and translating clades into ranks and taxa. In: Proceeding s of the \nFourth International Congress of Nematology, 8-13 June 2002, Tenerife, Spain. Brill, 633–\n653.  \nDeagle BE, Thomas AC, Shaffer AK, Trites AW, Jarman SN (2013) Quantifying sequence \nproportions in a DNA -based diet study using Ion Torrent amplicon sequ encing: which \ncounts count? Molecular Ecology Resources 13: 620 –633. https://doi.org/10.1111/1755-\n0998.12103  \nDeagle BE, Thomas AC, McInnes JC, Clarke LJ, Vesterinen EJ, Clare EL, Kartzinel TR, Eve son \nJP (2019) Counting with DNA in metabarcoding studies: How should we convert sequence \nreads to dietary data? Molecular Ecology 28: 391–406. https://doi.org/10.1111/mec.14734  \nDowle EJ, Pochon X, C Banks J, Shearer K, Wood SA (2016) Targeted gene enrichment and high-\nthroughput sequencing for environmental biomonitoring: a case study using freshwater \nmacroinvertebrates. Molecular Ecology Resources 16: 1240 –1254. \nhttps://doi.org/10.1111/1755-0998.12488 \nDyntaxa. Svensk taxonomisk databas https://www.dyntaxa.se/  \nEdgar RC (2010) Search and clustering orders of magnitude faster than BLAST. Bioinformatics \n26: 2460–2461. https://doi.org/10.1093/bioinformatics/btq461   \nEdgar RC (2016) UNOISE2: improved er ror-correction for Illumina 16S and ITS amplicon \nsequencing. BioRxiv: 81257. https://doi.org/10.1101/081257  \nElbrecht V, Leese F (2015) Can DNA -Based Ecosystem Assessments Quantify Species \nAbundance? Testin g Primer Bias and Biomass —Sequence Relationships with an \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nInnovative Metabarcoding Protocol. PLoS ONE 10: 1 –16. Available from: \nhttp://10.0.5.91/journal.pone.0130324  \nFerris H, Bongers T, De Goede RGM (2001) A framework for soil food web diagnostics: extension \nof the nematode faunal analysis concept. Applied Soil Ecology 18: 13 –29. \nhttps://doi.org/10.1016/S0929-1393(01)00152-4  \nFloyd R, A bebe E, Papert A, Blaxter M (2002) Molecular barcodes for soil nematode \nidentification. Molecular Ecology 11: 839 –850. https://doi.org/10.1046/j.1365-\n294X.2002.01485.x  \nGuardiola M, Uriz MJ, Taberlet P, Coissac E, Wangensteen OS, Turon X (2015) Deep-sea, deep-\nsequencing: metabarcoding extracellular DNA from sediments of marine canyons. PloS \nONE 10: e0139633. https://doi.org/10.1371/journal.pone.0139633  \nGuillou L, Bachar D, Audic S, Bass D, Berney C, Bittner L, Boutte C, Burgaud G, De Vargas C, \nDecelle J, Del Campo J, Dolan JR, Dunthorn M, Edvardsen B, Holzmann M, Kooistra \nWHCF, Lara E, Le Bescot N, Logares R, Mahé F, Massana R,  Montresor M, Morard R, \nNot F, Pawlowski J, Probert I, Sauvadet AL, Siano R, Stoeck T, Vaulot D, Zimmermann \nP, Christen R (2013) The Protist Ribosomal Reference database (PR2): a catalog of \nunicellular eukaryote small sub -unit rRNA sequences with curated t axonomy. Nucleic \nAcids Research 41: D597–D604. https://doi.org/10.1093/nar/gks1160  \nHebert PDN, Cywinska A, Ball SL, DeWaard JR (2003) Biological identifications through DNA \nbarcodes. Philosophical Tra nsactions of the Royal Society of London B: Biological \nSciences 270: 313–321. https://doi.org/10.1098/rspb.2002.2218  \nHolovachov O, Haenel Q, Bourlat SJ, Jondelius U (2017) Taxonomy assignment appro ach \ndetermines the efficiency of identification of OTUs in marine nematodes. Royal Society \nOpen Science 4: 170315. https://doi.org/10.1098/rsos.170315  \nHolterman M, van der Wurff A, van den Elsen S, van Megen H, Bongers T, Holovachov O, Bakker \nJ, Helder J (2006) Phylum -wide analysis of SSU rDNA reveals deep phylogenetic \nrelationships among nematodes and accelerated evolution toward crown clades. Molecular \nBiology and Evolution 23: 1792–1800. https://doi.org/10.1093/molbev/msl044  \nHolterman M, Rybarczyk K, van den Elsen S, van Megen H, Mooyman P, Santiago RP, Bongers \nT, Bakker J, Helder J (2008) A ribosomal DNA -based framework for the detection and \nquantification of stress -sensitive nematode families in terrestrial habitats. Molecular \nEcology Resources 8: 23–34. https://doi.org/10.1111/j.1471-8286.2007.01963.x  \nvan den Hoogen J, Geisen S, Routh D, Ferris H, Traunspurger W, Wardle DA, de Goede \nRGM, Adams BJ, Ahmad W, Andriuzzi WS et al. (2019) Soil nematode abundance and \nfunctional group composition at a global scale. Nature 572: 194 –198. \nhttps://doi.org/10.1038/s41586-019-1418-6  \nKawanobe M, Toyota K, Ritz K (2021) Development and application of a DNA metabarcoding \nmethod for comprehensive analysis of soil nematode communities. Applied Soil Ecology \n166: 103974. https://doi.org/10.1016/j.apsoil.2021.103974  \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nKatoh K, Standley  DM (2013) MAFFT multiple sequence alignment software version 7: \nimprovements in performance and usability. Molecular Biology and Evolution 30: 772 –\n780. https://doi.org/10.1093/molbev/mst010  \nKenmotsu H, Uchida K, Hirose Y, Eki T (2020) Taxonomic profiling of individual nematodes \nisolated from copse soils using deep amplicon sequencing of four distinct regions of the \n18S ribosomal RNA gene. PloS One 15: e0240336. \nhttps://doi.org/10.1371/journal.pone.0240336  \nKerfahi D, Tripathi BM, Porazinska DL, Park J, Go R, Adams JM (2016) Do tropical rain forest \nsoils have greater nematode diversity than High Arctic tundra? A metagenetic comparison \nof Malaysia and Svalbard. Global Ecology and Biogeography 25: 716 –728. \nhttps://doi.org/10.1111/geb.12448  \nKrehenwinkel H, Wolf M, Lim JY, Rominger AJ, Simison WB, Gillespie RG (2017) Estimating \nand mitigating amplification bias in qualitative and quantitative arthropod metabarcoding. \nScientific Reports 7: 1–12. https://doi.org/10.1038/s41598-017-17333-x  \nKylander ME, Bindler R, Martinez Cortizas A, Gallagher K, Mörth C-M, Rauch S (2013) A novel \ngeochemical approach to paleorecords of dust deposition and effective humidity: 8500 \nyears of peat accumulation at Store Mosse (the “Great Bog”), Sweden. Quaternary Science \nReviews 69: 69–82. \nLamb PD, Hunter  E, Pinnegar JK, Creer S, Davies RG, Taylor MI (2019) How quantitative is \nmetabarcoding: A meta -analytical approach. Molecular Ecology 28: 420 –430. \nhttps://doi.org/10.1111/mec.14920  \nMacheriotou L, Guili ni K, Bezerra TN, Tytgat B, Nguyen DT, Phuong Nguyen TX, Noppe F, \nArmenteros M, Boufahja F, Rigaux A, Vanreusel A, Derycke S (2019) Metabarcoding \nfree-living marine nematodes using curated 18S and CO1 reference sequence databases for \nspecies-level taxonomi c assignments. Ecology and Evolution 9: 1211 –1226. \nhttps://doi.org/10.1002/ece3.4814  \nMartínez Cortizas A, Sjöström JK, Ryberg EE, Kylander ME, Kaal J, L ópez-Costas O, Álvarez \nFernández N, Bindler R (2021) 9000 years of changes in peat organic matter composition \nin Store Mosse (Sweden) traced using FTIR -ATR. Boreas 50: 1161 –1178. \nhttps://doi.org/10.1111/bor.12527  \nMcMurdie PJ, Holmes S (2013) phyloseq : An R Package for Reproducible Interactive Analysis \nand Graphics of Microbiome Census Data. PLOS ONE 8: e61217. Available from: \nhttps://doi.org/10.1371/journal.pone.0061217  \nNaturvårdsverket (2015) Skötselplan för Store Mosse nationalpark. 81 pp. \nNeher DA (2001) Role of nematodes in soil health and their use as indicators. Journal of \nNematology 33: 161–168. \nPorazinska DL, Giblin-Davis RM, Faller L, Farmerie W, Kanzaki N, Morris K, Powers TO, Tucker \nAE, Sung W, Thomas WK (2009) Evaluating high-throughput sequencing as a method for \nmetagenomic analysis of nematode diversity. Molecular Ecology Resources 9: 1439–1450. \nhttps://doi.org/10.1111/j.1755-0998.2009.02611.x  \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nPorazinska DL, Sung WAY, Giblin -Davis RM, Thomas WK (2010a) Reproducibility of read \nnumbers in high-throughput sequencing analysis of nematode community composition and \nstructure. Molecular Ecology Resources 10: 666 –676. https://doi.org/10.1111/j.1755-\n0998.2009.02819.x  \nPorazinska DL, Giblin -Davis RM, Esquivel A, Powers TO, Sung WAY, Thomas WK (2010b) \nEcometagenetics confirm high tropical rainforest nema tode diversity. Molecular Ecology \n19: 5521–5530. https://doi.org/10.1111/j.1365-294x.2010.04891.x  \nPorazinska DL, Giblin -Davis RM, Powers TO, Thomas WK (2012) Nematode spatial and \necological patterns from tropical and temperate rainforests. PLoS One 7: e44641. \nhttps://doi.org/10.1371/journal.pone.0044641  \nPowers TO, Neher DA, Mullin P, Esquivel A, Giblin-Davis RM, Kanzaki N, Stock SP, Mora MM, \nUribe-Lorio L (2009) Tropical nematode diversity: vertical stratification of nematode \ncommunities in a Costa Rican humid lowland rainforest. Molecular Ecology 18: 985–996. \nhttps://doi.org/10.1111/j.1365-294x.2008.04075.x  \nPrice MN, Dehal PS, Arkin AP (2010) FastTree  2–approximately maximum-likelihood trees for \nlarge alignments. PloS One 5: e9490.  \nR Core Team (2021) R: A Language and Environment for Statistical Computing. Available from: \nhttps://www.r-project.org/  \nRStudio Team (2022) RStudio: Integrated Development Environment for R. Available from: \nhttp://www.rstudio.com/  \nRyberg EE, Väliranta M, Martinez -Cortizas A, Ehrlén J, Sjöström JK, Kylander ME (2022) \nPostglacial peatland vegetation succession in Store Mosse bog, south-central Sweden: An \nexploration of factors driving species change. Boreas 51: 651 –666. \nhttps://doi.org/10.1111/bor.12580  \nSapkota R, Nicolaisen M (2015) High-throughput sequencing of nematode communities from total \nsoil DNA extractions. BMC Ecology 15: 1–8. https://doi.org/10.1186/s12898-014-0034-4  \nSchenk J, Kleinbölting N, Traunspurger W (2020) Comparison of morphological, DNA barcoding, \nand metabarcoding characterizations of freshwater nematode communities. Ecology and \nEvolution 10: 2885–2899. https://doi.org/10.1002/ece3.6104  \nSikder MM, Veste rgård M, Sapkota R, Kyndt T, Nicolaisen M (2020) Evaluation of \nmetabarcoding primers for analysis of soil nematode communities. Diversity 12: 388. \nhttps://doi.org/10.3390/d12100388  \nSieriebriennikov B, Ferris H, de Goede RGM (2014) NINJA: an automated calculation system for \nnematode-based biological monitoring. European Journal of Soil Biology 61: 90 –93. \nhttps://doi.org/10.1016/j.ejsobi.2014.02.004  \nSohlenius B, Boström S (2001) Annual and long -term fluctuations of the nematode fauna in a \nSwedish Scots pine forest soil. Pedobiologia 45: 408 –429. https://doi.org/10.1078/0031-\n4056-00096  \nSohlenius B, Boström S, Ekebom A (1997) Metazoan microfauna in an ombrotrophic mire at \nAbisko, northern Sweden. European Journal of Soil Biology 33: 31–39.  \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101\n\nTaberlet P, Coissac E, Pompanon F, Brochmann C, Willerslev E (2012) Towards next-generation \nbiodiversity assessment using DNA metabarcoding. Molecular Ecology 21: 2045 –2050. \nhttps://doi.org/10.1111/j.1365-294X.2012.05470.x  \nThomas AC, Deagle BE, Eveson JP, Harsch CH, Trites AW (2016) Q uantitative DNA \nmetabarcoding: improved estimates of species proportional biomass using correction \nfactors derived from control material. Molecular Ecology Resources 16: 714 –726. \nhttps://doi.org/10.1111/1755-0998.12490  \nTreonis AM, Unangst SK, Kepler RM, Buyer JS, Cavigelli MA, Mirsky SB, Maul JE (2018) \nCharacterization of soil nematode communities in three cropping systems through \nmorphological and DNA metabarcoding approaches. Scientific Reports 8 : 1 –12. \nhttps://doi.org/10.1038/s41598-018-20366-5  \nVasselon V, Bouchez A, Rimet F, Jacquet S, Trobajo R, Corniquel M, Tapolczai K, Domaizon I \n(2018) Avoiding quantification bias in metabarcodin g: Application of a cell biovolume \ncorrection factor in diatom molecular biomonitoring. Methods in Ecology and Evolution \n9: 1060–1069. https://doi.org/10.1111/2041-210X.12960  \nWaeyenberge L, Sutter N de, Viaene N, Haegeman A (2019) New insights into nematode DNA -\nmetabarcoding as revealed by the characterization of artificial and spiked nematode \ncommunities. Diversity 11: 52. https://doi.org/10.3390/d11040052  \nWang Q, Garrity GM, Tiedje JM, Cole JR (2007) Naive Bayesian classifier for rapid assignment \nof rRNA sequences into the new bacterial taxonomy. Applied and Environmental \nMicrobiology 73: 5261–5267. https://doi.org/10.1128/aem.00062-07  \nWhitehead AG, Hemming JR (1965) A comparison of some quantitative methods of extracting \nsmall vermiform nematodes from soil. Annals of Applied Biology 55: 25 –38. \nhttps://doi.org/10.1111/j.1744-7348.1965.tb07864.x  \nYeates GW (1972) Nematoda of a Danish beech forest. I. Methods and general analysis. Oikos: \n178–189. https://doi.org/10.2307/3543403  \nYeates GW (1984) Variation in soil nematode diversity under pasture with soil and year. Soil \nBiology and Biochemistry 16: 95–102. https://doi.org/10.1016/0038-0717(84)90098-1  \nYeates GW, Bongers T, De Goede RG, Freckman DW, Georgieva SS (1993) Feeding habits in \nsoil nematode families and genera -an outline for soil ecologists. Journal of Nematology \n25: 315–331. \nZhou X, Li Y, Liu S, Yang Q, Su X, Zhou L, Tang M, Fu R, Li J, Huang Q (2013) Ultra -deep \nsequencing enables high -fidelity recovery of biodiversity for bulk arthropod samples \nwithout PCR amplification. Gigascience 2: 1–12. https://doi.org/10.1186/2047-217x-2-4  \n  \nAuthor-formatted, not peer-reviewed document posted on 01/02/2023. DOI:  https://doi.org/10.3897/arphapreprints.e101101","source_license":"CC-BY-4.0","license_restricted":false}