Is it possible to understand a book missing a quarter of the letters? Unveiling the belowground species richness of grasslands

https://doi.org/10.1016/j.agee.2021.107683Get rights and content

Highlights

  • Plant richness is higher below than aboveground in each type of studied grassland.

  • Barcoding shows ca. 65% more species belowground if compared to observer survey.

  • Applying metabarcoding for both above and belowground reduces this proportion.

  • Difference between below- and aboveground richness drops with increased soil fertility.

  • Management intensity influences aboveground species richness, but not belowground.

Abstract

Knowledge of species richness patterns in plant communities is biased toward its hidden belowground part and is derived only from the part visible aboveground. Studies on the relationship of the above- to belowground parts of plant communities suffer from the lack of a consistent and uniform approach to assess their taxonomic composition, particularly in species-rich ecosystems. This study described the proportion between above- and belowground parts of vegetation in three grassland types along with the environmental factors that affect them, using eDNA metabarcoding and an observer survey. The internal transcribed spacer 2 (ITS2) region was used to analyze the total plant richness in the soil and the aboveground “green” part of plant communities. Considering all samples, eDNA metabarcoding successfully assigned about 93% of taxa on the species level. Our findings support the notion that metabarcoding analysis of the belowground plant community revealed up to 66% higher species richness than was identified above the ground by a conventional vegetation survey. However, this difference decreases to about 25% when an eDNA-derived taxonomic assignment was applied for the aboveground part of phytocoenosis. We also found that the difference between the below- and aboveground parts of the plant community decreased with increased soil fertility. Finally, the management intensity was found to significantly influence species richness only in the aboveground part of vegetation with the belowground part of grassland remaining unaffected. Overall, since DNA-based metabarcoding and traditional vegetation surveys have limitations, their complementary use is recommended to obtain the most reliable vegetation description. However, it should be considered that metabarcoding analysis is destructive and may not be applicable in protected or private areas or on permanent research plots.

Introduction

Despite significant progress in exploring belowground plant richness, knowledge about the patterns of species coexistence, the structure of belowground communities and their response to environmental variables are still far from sufficiently explored (Matesanz et al., 2019, Wilson, 2014). Due to obvious observational constraints, even the simple count of taxa in a community and the examination of its composition and abundance suffer considerable inexactness because of flaws in identifying of roots based solely on morphological traits (Silva and Rego, 2003). Recently, several cutting-edge studies have shown that depending on the geographical position, belowground plant richness can be from 50% to 100% higher than aboveground (Pärtel et al., 2012, Träger et al., 2019). This proportion is caused by heterogeneous nature of soil in space and time, the storage ability of soil, and symbioses with many microorganisms that promote belowground plant species richness (Pärtel et al., 2012). Additionally, the fast development of metabarcoding methods has enabled much broader, cheaper and more accurate exploration of belowground richness and abundance (Li et al., 2017, Matesanz et al., 2019, Mommer et al., 2011). However, although significant progress in exploring of belowground species composition was done, most ecosystems have not yet been studied. Although studies conducted so far (Hiiesalu et al., 2012, Li et al., 2017, Matesanz et al., 2019, Träger et al., 2019) have brought crucial knowledge showing the general patterns of the proportions of above- to belowground plant richness, still many questions remain unsolved. The knowledge on the patterns of the belowground part of plant communities needs to be supplemented by research that concerns various vegetation types with different management regimes, different sample sizes, and new approaches in metabarcoding and sequencing. As a result we must obtain a comprehensive and more complete picture of the relationships between above- and belowground plant composition in various vegetation types and understand all patterns and gradients on which the researched phytocoenoses depend.

When analyzing the proportion of below- to aboveground vegetation, it is crucial to include complementary methods for species identification. The observer error, inherently involved in all research, can reach 20–30% in plant richness surveys (Archaux et al., 2006). Misidentifications occur, and there is considerable variation between the observers who sample the vegetation (Morrison, 2016). In some cases, low detectability of particular species makes it impossible to find (Moore et al., 2011) due to insufficient observer experience, high average cover, density and height of the stand, its phenological dynamics, and species richness. In dense, high, lush, species-rich and seasonally extended vegetation, the difference between the detected and true number of plants increases (Dexter et al., 2010, Moore et al., 2011). Additionally, several species that molecular analyses can easily detect are hardly noticeable by the eye. This applies to individuals that occur with hardly identifiable organs, such as small cotyledons, upper parts of bulbs that extend aboveground, and often pale parts of decumbent, procumbent, or prostrate stolons or runners (sarments) and short or long-lived epigenous rhizomes (Klimešová and Klimeš, 2007). Despite thorough research, aboveground surveys can overlook immature, tiny seedlings or young, often white or colorless coleoptiles of grasses. The DNA metabarcoding of the aboveground part of phytocoenoses may also be affected by various errors. When applying molecular techniques, fragments of withered plants, residues from the previous season, and fragments of plants transported from adjacent areas adhered or tangled with the sample, and difficult to remove when cleaning the material, are identified. These shortcomings lead to the conclusion that methodological complementarity should be maintained, and the same method should be applied when investigating the above- and belowground parts of the plant community.

The recent development of environmental DNA metabarcoding (eDNA) due to considerable cost reduction and improved bioinformatic analyses has improved the efficiency and reliability of mixed-species samples (Herben et al., 2017, Hiiesalu et al., 2012, Li et al., 2017). A growing number of researchers apply DNA metabarcoding for plant species detection in a given habitat or environmental sample (Banchi et al., 2020, Coghlan et al., 2020). Most studies have been based on the analysis of different regions of the chloroplast DNA, trnL (UAA) intron, encoded RuBisCo large subunit (rbcL) gene to find the species composition of different groups of plants in the soil ecosystem (Hiiesalu et al., 2012, Pärtel et al., 2012, Soininen et al., 2009, Sønstebø et al., 2010). More attention is focused on the internal transcribed spacer (ITS) of the nuclear ribosomal cistron (18S-5.8S-26S) for metabarcoding purposes, mainly due to its high taxonomic resolution (Bell et al., 2016; Cheng et al., 2016). The ca 5k plant species indicated the correct taxonomic identification at the species and genus levels as approximately 91.5% and 99.8% for the ITS2 region (Chen et al., 2010). It is also recommended for taxonomic assignment in the Poaceae and Asteraceae families (Brennan et al., 2018, De Barba et al., 2014), which dominate grassland communities (Chytrý, 2007). As one of the most widely used DNA fragments in plant molecular systematics at the generic and species levels (Staggemeier et al., 2015, Yuan et al., 2015), ITS sequences are often a major part of large public databases facilitating the identification of unknowns (e.g., GenBank). Moreover, using the ITS for species-level identification is especially advantageous and more balanced than other plastid regions because green tissues of leaves, cotyledons, and stems contain more plastome copies per organelle than non-green tissues, such as root tissues from belowground (Liere and Börner, 2013, Ma and Li, 2015).

Despite obvious advantages of metabarcoding, many factors can considerably influence and alter the correspondence between the percentage of reads retrieved and the species’ abundance in the field (Deiner et al., 2017, Gloor et al., 2017, Porter and Hajibabaei, 2018). Analyses of samples in ecological studies are sensitive to the technical and methodological factors, such as cross-sample contamination and polymerase chain reaction (PCR) errors (Calderón-Sanou et al., 2020). It is estimated that more than 70% of the singletons in raw metabarcoding include errors produced during DNA extraction and sequencing (Brown et al., 2015). Additionally, the PCR used to prepare the DNA library may introduce quantitative biases due to primer-template mismatches and differences in amplicon length (Brodin et al., 2013, Parada et al., 2016, Tedersoo et al., 2015). This inhibition can affect species detectability and reproducibility. Therefore, removing contaminants that inhibit PCR reaction remains the key issue for estimating and identifying species in metabarcoding research (Rucińska et al., 2021, Uchii et al., 2019). To improve its performance, other possible limitations of metabarcoding were also considered (PCR bias and taxonomic assignment of amplicon sequencing variants).

The novelty of our approach was the design of the experiment that included different types of species-rich meadows managed in different way. Grasslands are extremely rich in plant species; however, diversity patterns of vascular plants vary across vegetation types (Biurrun et al., 2021). The highest richness at plots below 50 m2 was found on semi-natural, oligo- to mesotrophic, temperate grasslands (Wilson et al., 2012). The biodiversity of grassland depends to a large extent on management intensity. Central European grasslands owe large species richness to extensive use. Intensification of management can dramatically decrease the diversity and richness of grassland organisms (Heuss et al., 2019). On the other hand, reduced mowing intensity or abandonment can also lead to species loss (Milberg et al., 2017). Patterns of species composition of grasslands in different vegetation types and under the influence of various management regimes are relatively well known. An additional factor influencing the changes in species composition is habitat fertility, however with ambiguous effects on the belowground and aboveground parts of the phytocoenosis. Species richness in the aboveground part of the grassland vegetation decreases with increasing fertility (Mittelbach et al., 2001). Additionally, Hiiesalu et al. (2012) reported that the belowground richness of graminoid vegetation was not related to soil fertility.

In this study, we addressed the nature of the relationship between above- and belowground plant species richness and composition in Central European grasslands. Our novel approach was to sample three types of grasslands (wet, moderately wet, and dry) and three types of management intensity in three subsequent seasons (spring, summer, and autumn). Our primary hypotheses were the following: (i) the relation of above- and belowground plant richness depends on the methods applied for identifying species due to observer errors during field studies; (ii) the difference between below- and aboveground species composition is lower when examined based on only molecular identification of plants compared with the proportion obtained from visual census aboveground; (iii) the relation between below- and aboveground parts of phytocoenoses depends on the season of sampling and type of vegetation; and (iv) management intensity influences phytocoenoses aboveground more than belowground.

Section snippets

Materials and methods

To describe the relationship between above- and belowground plant richness in Central European grasslands, eDNA metabarcoding analysis was applied with the commonly used MiSeq Illumina sequencing. An eDNA metabarcoding protocol was established in this study to minimize bias concerning the accurate taxonomic representation of species richness in the following steps (i) maximizing the quantity and quality of eDNA free of contamination obtained from a large volume of plant dry biomass, (ii)

The reference database

From the 187 plant specimens sequenced to generate a reference database for the ITS2 region, 186 sequences were recovered, corresponding to 186 species and lacking only Equisetum pratense, probably due to incorrect starter adjustment. Some species shared identical sequences, especially those representing the genus Carex (Fig. 1). The length of sequences varied from 260 bp for Arabis hirsuta to 492 bp for Scirpus sylvaticus (Table S1).

Mock community libraries

The 21 libraries prepared from the mixed DNA of 13 plant

Species richness is higher below- than aboveground in different Central European grasslands

Similar to previous studies, a higher belowground species richness was confirmed for Central European grasslands (Hiiesalu et al., 2012, Pärtel et al., 2012). This rule holds for all types of habitats, in all types of management, at any time of the growing season, and for two sample sizes. We detected less total species richness on the aboveground parts of vegetation than belowground by 6–19% on 0.9 m2 and 13–37% on 0.1 m2. The difference between the above- and belowground parts increases if we

Conclusions

Conventional vegetation surveys conducted by field botanists are still the method of choice for aboveground plant species identification. However, as the visible is not always seen, and as they are prone to observer errors (Morrison, 2016) with a tendency to underestimate the total plant richness, such surveys should be supported by DNA-based methods of plant identification applied to the belowground part of vegetation and the part above the ground, particularly when the difference between

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This study was funded by the National Science Centre, Poland, project No. 2017/25/B/NZ8/00572 and partially by project No. 2019/35/B/NZ8/03358.

References (103)

  • M. Pärtel et al.

    Dark diversity: shedding light on absent species

    Trends Ecol. Evol.

    (2011)
  • J. Weiner

    Asymmetric competition in plant populations

    Trends Ecol. Evol.

    (1990)
  • H. Ando et al.

    Evaluation of plant contamination in metabarcoding diet analysis of a herbivore

    Sci. Rep.

    (2018)
  • Andrews, S., 2010. FastQC: a quality control tool for high throughput sequence...
  • F. Archaux et al.

    Effects of sampling time, species richness and observer on the exhaustiveness of plant censuses

    J. Veg. Sci.

    (2006)
  • D. Bates et al.

    Fitting linear mixed-effects models using lme4

    J. Stat. Softw.

    (2015)
  • G.R. Beard et al.

    The value of consistent methodology in long-term environmental monitoring

    Environ. Monit. Assess.

    (1999)
  • K.L. Bell et al.

    Pollen DNA barcoding: current applications and future prospects

    Genome

    (2016)
  • G. Blume-Werry et al.

    The hidden season: growing season is 50% longer below than above ground along an arctic elevation gradient

    New Phytol.

    (2016)
  • N.A. Bokulich et al.

    Quality-filtering vastly improves diversity estimates from Illumina amplicon sequencing

    Nat. Methods

    (2013)
  • A.M. Bolger et al.

    Trimmomatic: a flexible trimmer for Illumina sequence data

    Bioinformatics

    (2014)
  • E. Bolyen et al.

    Reproducible, interactive, scalable and extensible microbiome data science using QIIME 2

    Nat. Biotechnol.

    (2019)
  • G.-J. Brandon-Mong et al.

    DNA metabarcoding of insects and allies: an evaluation of primers and pipelines

    Bull. Entomol. Res.

    (2015)
  • G.L. Brennan et al.

    Temperate grass allergy season defined by spatio-temporal shifts in airborne pollen communities

    bioRxiv

    (2018)
  • J. Brodin et al.

    PCR-induced transitions are the major source of error in cleaned ultra-deep pyrosequencing data

    PLoS One

    (2013)
  • I. Biurrun et al.

    Benchmarking plant diversity of Palaearctic grasslands and other open habitats

    J. Veg. Sci.

    (2021)
  • I. Calderón-Sanou et al.

    From environmental DNA sequences to ecological conclusions: how strong is the influence of methodological choices?

    J. Biogeogr.

    (2020)
  • B.J. Callahan et al.

    DADA2: high-resolution sample inference from Illumina amplicon data

    Nat. Methods

    (2016)
  • A. Chao et al.

    Rarefaction and extrapolation with Hill numbers: a framework for sampling and estimation in species diversity studies

    Ecol. Monogr.

    (2014)
  • Chao, A., Ma, K.H., Hsieh, T.C., Chiu, C.-H., 2016. SpadeR: Species-Richness Prediction and Diversity Estimation with...
  • S. Chen et al.

    Validation of the ITS2 region as a novel DNA barcode for identifying medicinal plant species

    PLoS One

    (2010)
  • T. Cheng et al.

    Barcoding the kingdom Plantae: new PCR primers for ITS regions of plants with improved universality and specificity

    Mol. Ecol. Resour.

    (2016)
  • L.J. Clarke et al.

    Environmental metabarcodes for insects: in silico PCR reveals potential for taxonomic bias

    Mol. Ecol. Resour.

    (2014)
  • S.A. Coghlan et al.

    Development of an environmental DNA metabarcoding assay for aquatic vascular plant communities

    Environ. DNA

    (2020)
  • J. Dabney et al.

    Length and GC-biases during sequencing library amplification: a comparison of various polymerase-buffer systems with ancient and modern DNA sequencing libraries

    Biotechniques

    (2012)
  • M. De Barba et al.

    DNA metabarcoding multiplexing and validation of data accuracy for diet assessment: application to omnivorous diet

    Mol. Ecol. Resour.

    (2014)
  • E.G. De La Riva et al.

    The importance of functional diversity in the stability of Mediterranean shrubland communities after the impact of extreme climatic events

    J. Plant Ecol.

    (2017)
  • K. Deiner et al.

    Environmental DNA metabarcoding: Transforming how we survey animal and plant communities

    Mol. Ecol.

    (2017)
  • K.G. Dexter et al.

    Using DNA to assess errors in tropical tree identifications: How often are ecologists wrong and when does it matter?

    Ecol. Monogr.

    (2010)
  • M.E. Edwards et al.

    Metabarcoding of modern soil DNA gives a highly local vegetation signal in Svalbard tundra

    Holocene

    (2018)
  • V. Elbrecht et al.

    Validation and development of COI metabarcoding primers for freshwater macroinvertebrate bioassessment

    Front. Environ. Sci.

    (2017)
  • V. Elbrecht et al.

    Can DNA-based ecosystem assessments quantify species abundance? Testing primer bias and biomass—Sequence relationships with an innovative metabarcoding protocol

    PLoS One

    (2015)
  • P. Ewels et al.

    MultiQC: summarize analysis results for multiple tools and samples in a single report

    Bioinformatics

    (2016)
  • D.A. Frank et al.

    Fine-scale belowground species associations in temperate grassland

    Mol. Ecol.

    (2015)
  • G.T. Freschet et al.

    Integrated plant phenotypic responses to contrasting above- and below-ground resources: key roles of specific leaf area and root mass fraction

    New Phytol.

    (2015)
  • S. Getzin et al.

    Heterogeneity influences spatial patterns and demographics in forest stands

    J. Ecol.

    (2008)
  • S. Goodwin et al.

    Coming of age: ten years of next-generation sequencing technologies

    Nat. Rev. Genet.

    (2016)
  • G.B. Gloor et al.

    Microbiome datasets are compositional: And this is not optional

    Front. Microbiol.

    (2017)
  • M. Hajibabaei et al.

    Environmental Barcoding: a next-generation sequencing approach for biomonitoring applications using river benthos

    PLoS One

    (2011)
  • Cited by (4)

    • Comparison of the predictive ability of spectral indices for commonly used species diversity indices and Hill numbers in wetlands

      2022, Ecological Indicators
      Citation Excerpt :

      The Hill numbers index assigns different weights to dominant and rare species in the community by setting different q values, which can achieve a continuous measure of community species diversity (Hill, 1973). After Hill et al. (1973) proposed the Hill numbers index in 1973 (Hill, 1973), Chao et al. (2014) further expanded and improved it (Chao et al., 2014), which had been widely used in other fields of ecology (Kifle et al., 2022; Rucinska et al., 2022; Siegenthaler et al., 2022), but have not been used in the study of the spectral-species diversity. Based on the above research background, we hypothesized that (1) the predictive ability for species diversity of using NDVIMEAN alone is limited and the combination of NDVIMEAN and NDVISD can significantly improve the predictive ability of species diversity. (

    View full text