Elsevier

Flora

Volume 254, May 2019, Pages 161-172
Flora

On the relevance of intraspecific trait variability—A synthesis of 56 dry grassland sites across Europe

https://doi.org/10.1016/j.flora.2019.03.002Get rights and content

Highlights

  • Intraspecific exceeds between species trait variability for most traits.

  • Between population exceeds within population trait variability.

  • The traits of different species vary mainly consistently between sites.

  • Climate factors can help explaining between population trait variability.

  • The proportion of intraspecific trait variability diminishes in the n-dimensional trait space.

Abstract

The intraspecific plant trait variability (ITV) is key for many ecological processes, but large-scale analysis of co-occurring species are rare. Here we studied ITV of 14 plant traits for five co-occurring species across 56 European dry grassland sites – large parts of the species distribution ranges. We evaluated variation in ITV relative to between species trait variability (BTV) with a particular emphasis on how within versus between population variability contributes to ITV. We performed this analysis trait by trait (univariate) and by considering hypervolumes in the multi-variate trait space. We also tested whether climatic variables can be used to predict between population variation in traits.

For the trait by trait analysis, for 9 out of 14 traits ITV exceeded BTV, which suggests an extraordinary large role of ITV. However, when considering all traits jointly the hypervolume occupied by ITV was only 23% of overall trait variability.

We found comparatively small effects of within population variability in the uni- as well as in the multivariate analysis decreasing rapidly with increasing number of considered traits. The dominance of between population variability suggests that ITV is mainly driven by environmental effects rather than local biotic interactions and microsite effects. ITV of leaf chemical traits was related to precipitation and growing degree days until sampling

ITV can be substantial compared to BTV, in particular when considering single traits, such that ITV should be considered in trait-based research. Since the importance of ITV appears to decrease when considering multiple traits jointly, using species mean trait values is more appropriate for multi-variate trait analysis. Because ITV varied mainly between populations rather than within populations, and was related to climatic conditions, we suggest that ITV could be accounted for in future trait based research by including environmental covariates in hierarchical models of trait variation

Introduction

Recent research has identified that plant traits of different species co-vary with other traits (Reich, 2014; Reich et al., 1999; Wright et al., 2004), with environmental conditions (Ivanova et al., 2019; Tautenhahn et al., 2008), land-use (Römermann et al., 2009), phenology (König et al., 2018), and species distributions (Stahl et al., 2014) suggesting that traits reflect the strategies plants use to cope with the environment. Furthermore, plant traits have been used to predict feedbacks of vegetation on ecosystem processes (Musavi et al., 2015; Reichstein et al., 2014) and to inform ecosystem models (Kattge et al., 2009). Such analyses pool measured trait values into species mean trait values, while the intraspecific trait variability (in the following called ITV) is generally ignored (Pérez-Harguindeguy et al., 2013). However, traits vary not only between species (in the following called BTV) but also within species (Schlichting, 1986). ITV can be due to phenotypic plasticity within one genotype or genotypic variation within one species (Franks et al., 2014; Fridley and Grime, 2015; Leimu and Fischer, 2008; Nicotra et al., 2010).

The basic assumption behind ignoring ITV in trait based research is that ITV is smaller than BTV (Kattge et al., 2011; Westoby et al., 2002). Indeed, many studies have shown that BTV exceeds ITV (Albert et al., 2010; Cornelissen et al., 2003; Kattge et al., 2011; Roche et al., 2004; Steinbeiss et al., 2008; Garnier et al., 2019), yet other studies have found ITV to be substantial relative to BTV (Jung et al., 2010; Mitchell and Bakker, 2014). However, none of these studies analyzed ITV over large spatial extents covering the species distribution ranges for co-occurring species and therefore likely underestimate ITV. It has further been shown that also BTV increases with spatial extent (Siefert et al., 2015), suggesting that the relative contribution of ITV is unimportant in large scale studies. On the other hand, the effects of changing site conditions along a large gradient also shape ITV, which implies that the relative contribution of ITV vs. BTV to overall trait variability depends on the strength of site effects on ITV relative to the strength of site effects on species turnover (Leps et al., 2011; Rosbakh et al., 2015).

The primary source of ITV is assumed to be between population variation (in the following called ITVbp; Albert et al., 2010) and thus ITV has mainly been studied along spatial, environmental, or management gradients (Auger and Shipley, 2013; Bucher et al., 2016; Fajardo and Piper, 2011; Hulshof et al., 2013; Soper Gorden et al., 2016; Sun et al., 2015). The literature on how within population variability (in the following called ITVwp) contributes to ITV is limited. Phenology (Römermann et al., 2016), ontogeny (Coleman et al., 1994), micro-site, or biotic effects (Le Bagousse-Pinguet et al., 2015) have been identified as sources of ITVwp. The relative importance of within versus between population ITV is highly relevant for trait based research because ITVbp may be predictable from e.g. regional effects whereas ITVwp seems to be controlled by local effects, which are less predictable.The variability of traits, whether between species, between populations, or within populations, has mostly been studied on a trait by trait basis. However, a multivariate perspective that considers several traits can improve the understanding of functional adaptations of plants to the environment (Albert et al., 2010; Díaz et al., 2016). The n-dimensional trait space which can be occupied by a species can be interpreted as the species functional niche (Boucher et al., 2013), similar to the original niche concept (Hutchinson, 1957) of an n-dimensional space of environmental conditions in which a species needs to complete its life cycle. The mean trait values of a species can determine its niche position along gradients, and intraspecific trait variability determines its niche breadth (Stahl et al., 2014; Violle and Jiang, 2009). Such n-dimensional trait niches can be adequately described using hypervolumes (Blonder, 2018), which therefore provide a measure of functional diversity (Lamanna et al., 2014). Despite the conceptual appeal, the relative contributions of BTV, ITVbp, and ITVwp in the multivariate trait space have yet to be quantified.

In this study, we investigate the degree of ITV relative to BTV and the degree of ITVwp relative to ITVbp for five co-occurring dry grassland species that were sampled over a wide geographic gradient capturing most of their environmental space in Central Europe. We analyse the data using both a univariate trait by trait perspective and a multivariate perspective that examines all traits jointly in the n-dimensional trait space. We also investigate the potential of predicting ITVbp from climate variables. Specifically, we address the following questions:

  • 1)

    What is the magnitude of ITV relative to BTV over a large spatial extent?

  • 2)

    Is ITV primarily due to ITVwp or ITVbp?

  • 3)

    Is between population variability consistent across species or species specific and related to climate?

  • 4)

    Do univariate and multivariate trait perspectives yield consistent patterns?

We will address these questions for five co-occurring species (Plantago media L., Campanula glomerata L., Trifolium montanum L., Centaurea jacea L., and Salvia pratensis L.) from 56 dry grassland sites across Europe. This study is unique with respect to the sampling of most of the species co-occurrence range which facilitates unbiased estimates on the relative importance of ITV versus BTV. European dry grasslands represent suitable study systems as they have usually comparable management and land-use histories (Poschlod and WallisDeVries, 2002). Dry grassland species need to cope with nutrient-poor and dry conditions and accordingly show morphological and eco-physiological adaptations to the stressful environment. Differences in trait values between populations of the same species have been observed in these habitats (Lauterbach et al., 2013), making them interesting study systems in particular in the context of climate and / or land use changes.

Section snippets

Study sites and species

We studied the trait variability of the five focal species Plantago media L., Campanula glomerata L., Trifolium montanum L., Centaurea jacea L., and Salvia pratensis L. on 56 dry grassland sites across Europe (Fig. 1). These species are endemic to dry grassland habitats (Korneck, 1996; Oberdorfer et al., 1978). Our five study species are all hemicryptophytes and typically co-occur over wide gradients in Europe (Kalwij et al., 2014). The study sites have similar land-use histories and soil

Univariate analysis of trait variability

The boxplots of trait distributions per species across all sites in Fig. 3 shows that ITV is large. The ANOVA analysis revealed that overall trait variability of all considered traits depends significantly on “species”, “site effects”, and their interaction (Table 2). These three factors together accounted for between 55% (plant width) and 85% (stomata density) of overall trait variability (Fig. 4). The remaining unexplained variability ηresiduals2 represents ITVwp which can be attributed to

Intraspecific versus between species trait variability

In the univariate analyses the measured trait variability was dominated by intraspecific trait variability (ITV) relative to between species trait variability (BTV) for 9 out of 14 traits. For some traits such as specific leaf area ITV accounts for 92% of the overall trait variability. Even when considering groups of traits in an n-dimensional space ITV exceeded BTV except for the group of stomata traits (ϑITV = 44%). The dominance of ITV is surprising, given that previous studies have

found

Implications for trait-based research

We demonstrated that the assumption of negligible ITV is inappropriate in our large-scale synthesis of dry grasslands when analysing trait variability trait by trait. For the five observed dry grassland species ITV exceeds BTV for the majority of traits. However, we could also show that in the multivariate trait space ITV is smaller than BTV if a sufficiently large number of traits (∼4) are considered jointly. Overall the effect of ITV relative to BTV decreases rapidly with increasing trait

Research data for this article

Due to ongoing PhD theses based on this data-set the raw data remain confidential and cannot be shared within this publication. However, the data will be available in future publications.

Data not available / the data that has been used is confidential.

Acknowledgements

We are thankful to all cooperating scientists for support in field work and site selection in the different countries. Anna Golinko, Andrea Bauer, Franziska Pfeiffer, Verena Wommer and Dörte Bachmann helped with trait measurements in the field. We thank Jürgen Dengler and the European dry grassland group (EDDG) for looking for collaborators, finding sites and Ute Jandt to help defining adequate focal species. The leaf nutrient analyses were cooperatively performed at the laboratory of the Dry

References (72)

  • B. Blonder et al.

    The n-dimensional hypervolume

    Glob. Ecol. Biogeogr.

    (2014)
  • F.C. Boucher et al.

    Unravelling the architecture of functional variability in wild populations of Polygonum viviparum L

    Funct. Ecol.

    (2013)
  • S.F. Bucher et al.

    Inter- and intraspecific variation in stomatal pore area index along elevational gradients and its relation to leaf functional traits

    Plant Ecol.

    (2016)
  • S. Casson et al.

    Influence of environmental factors on stomatal development

    New Phytol.

    (2008)
  • J.S. Coleman et al.

    Interpreting phenotypic variation in plants

    Trends Ecol. Evol.

    (1994)
  • J.H.C. Cornelissen et al.

    A handbook of protocols for standardised and easy measurement of plant functional traits worldwide

    Aust. J. Bot.

    (2003)
  • P. De Frenne et al.

    Latitudinal gradients as natural laboratories to infer species’ responses to temperature

    J. Ecol.

    (2013)
  • S. Díaz et al.

    The global spectrum of plant form and function

    Nature

    (2016)
  • A. Fajardo et al.

    Intraspecific trait variation and covariation in a widespread tree species (Nothofagus pumilio) in southern Chile

    New Phytol.

    (2011)
  • S.J. Franks et al.

    Evolutionary and plastic responses to climate change in terrestrial plant populations

    Evol. Appl.

    (2014)
  • J.D. Fridley et al.

    Community and ecosystem effects of intraspecific genetic diversity in grassland microcosms of varying species diversity

    Ecology

    (2015)
  • E. Garnier et al.

    Inter- and intra- specific trait shifts among sites differing in drought

    Flora

    (2019)
  • I. Harris et al.

    Updated high-resolution grids of monthly climatic observations – the CRU TS3.10 dataset

    Int. J. Climatol.

    (2014)
  • K.W. Hilu et al.

    Convenient method for studying grass leaf epidermis

    Taxon

    (1984)
  • N. Holland et al.

    Stomatal length correlates with elevation of growth in four temperate species

    J. Sustain. For.

    (2009)
  • C.M. Hulshof et al.

    Intra-specific and inter-specific variation in specific leaf area reveal the importance of abiotic and biotic drivers of species diversity across elevation and latitude

    J. Veg. Sci.

    (2013)
  • E. Hutchinson

    Concluding remarks and future work

    Symp. Quant. Biol.

    (1957)
  • L.A. Ivanova et al.

    Leaf traits of C3- and C4-plants indicating climatic adaptation along a latitudinal gradient in Southern Siberia and Mongolia

    Flora

    (2019)
  • V. Jung et al.

    Intraspecific variability and trait-based community assembly

    J. Ecol.

    (2010)
  • R.R. Junker et al.

    Dynamic range boxes – a robust nonparametric approach to quantify size and overlap of n-dimensional hypervolumes

    Methods Ecol. Evol.

    (2016)
  • J.M. Kalwij et al.

    Spatially- explicit estimation of geographical representation in large-scale species distribution datasets

    PLoS One

    (2014)
  • J. Kattge et al.

    Quantifying photosynthetic capacity and its relationship to leaf nitrogen content for global-scale terrestrial biosphere models

    Glob. Change Biol.

    (2009)
  • J. Kattge et al.

    TRY - a global database of plant traits

    Glob. Change Biol.

    (2011)
  • E. Kazakou et al.

    Are trait-based species rankings consistent across data sets and spatial scales?

    J. Veg. Sci.

    (2014)
  • P. König et al.

    Advances in flowering phenology across the Northern Hemisphere are explained by functional traits

    Glob. Ecol. Biogeogr.

    (2018)
  • D. Korneck

    Warum verarmt unsere Flora? Auswertung der Roten Liste der Farn- und Blütenpflanzen Deutschlands

    Schriftenreihe für Vegetationskunde Bundesamt für Naturschutz

    (1996)
  • Cited by (20)

    • Reprint of: Drivers of within-tree leaf trait variation in a tropical planted forest varying in tree species richness

      2021, Basic and Applied Ecology
      Citation Excerpt :

      These differences in light availability influence leaf traits: for example, LMA and leaf C/N ratio are known to increase with increasing light availability (Böhnke & Bruelheide, 2013), while leaf N decreases (Zhao & Oosterhuis, 1998). Regarding biotic interactions, the prevailing pattern described in the literature is that community trait space increases in species mixtures compared to monocultures (Siefert et al., 2015; Raffard, Santoul, Cucherousset & Blanchet, 2019; Tautenhahn, Grün-Wenzel, Jung, Higgins & Römermann, 2019). In contrast, the relationship between species richness of a community and within-species trait variation (the species niche breadth) has been discussed controversially.

    • Intraspecific trait variability of a typical tree species of riverine forests in southern Brazil

      2021, Flora: Morphology, Distribution, Functional Ecology of Plants
      Citation Excerpt :

      Considering the variation within and between trees, Messier et al. (2017) observed greater variability of LA in these levels than for plot and site scales. The larger variation of LA at the intra-population scale is related to biotic interactions and responses to heterogeneous conditions of resource availability (Moreira et al., 2012; Tautenhahn et al., 2019). In this study, leaf area shows to suffer positive effects of the difference between mean canopy height and mean height of G. klotzschiana, i.e. this difference between mean heights increases the shading and, consequently, the LA increases in response to the decrease in light availability.

    View all citing articles on Scopus

    This article is part of a special issue entitled: “Functional traits explaining plant responses to past and future climate changes” published at the journal Flora 254C, 2019.

    1

    Authors contributed equally to this work.

    View full text