Skip to main content
Dryad

Data from: Measuring leaf and root functional traits uncovers multidimensionality of plant responses to arbuscular mycorrhizal fungi

Cite this dataset

Stahlhut, Katherine et al. (2024). Data from: Measuring leaf and root functional traits uncovers multidimensionality of plant responses to arbuscular mycorrhizal fungi [Dataset]. Dryad. https://doi.org/10.5061/dryad.2z34tmpv1

Abstract

Premise of the study While many studies have measured the aboveground responses of plants to mycorrhizal fungi at a single time point, little is known about how plants respond belowground or across time to mycorrhizal symbiosis. By measuring belowground responses as well as growth over time in many plant species, we create a more complete picture of how mycorrhizal fungi benefit their hosts. Methods We grew 26 prairie plant species with and without mycorrhizal fungi and measured fourteen functional traits measuring above and belowground tissue quality and quantity responses and changes in resource allocation. We used function-value trait (FVT) modeling to characterize changes in species growth rate when colonized. Key results While aboveground biomass responses were positive, the response of traits belowground were much more variable. Changes in aboveground biomass accounted for 60.8% of the variation in mycorrhizal responses, supporting the use of aboveground biomass response as the primary response trait. Responses belowground were not associated with aboveground responses and accounted for 18.3% of the variation. Growth responses over time were highly variable across species. Interestingly, none of the measured responses were phylogenetically conserved. Conclusions Mycorrhizal fungi increase plant growth in most scenarios, but the effects of these fungi belowground and across time are more complicated. This study highlights how differences in plant allocation priorities might affect how they utilize the benefits from mycorrhizal fungi. Identifying and characterizing these differences is a key step to understanding the effects of mycorrhizal mutualisms on whole plant physiology. 

README: Measuring leaf and root functional traits uncovers multidimensionality of plant responses to arbuscular mycorrhizal fungi

https://doi.org/10.5061/dryad.2z34tmpv1

The data and code presented here contains all of the necessary information to complete analyses in the paper "Measuring leaf and root functional traits uncovers multidimensionality of plant responses to arbuscular mycorrhizal fungi" published in American Journal of Botany. The includes the following five parts:

  1. Calculating Plant Functional Traits
  2. Calculating Response Ratios
  3. Calculating Phylogenetic Signal
  4. Creating Trait PCAs
  5. Conducting Function-Value Trait Analysis

Description of the data and file structure

The following code files are included in this submission:

  • CalculatePFTs.R: Calculates the plant traits used in this analysis, using the raw data file* RawData.csv*. This creates a data frame with only the focal traits called* RawData1.csv*, which is used in* ResponseRatios.R*
  • ResponseRatios.R: Calculates the response ratios of each focal trait, using the file RawData1.csv. This creates a new data frame with the response ratios named TraitsMeans.csv. It also creates Response_boxplots.tiff, which is Figure 1 in the paper.
  • Phylosig.R: Calculates Pagel’s λ and Blomberg’s for each response ratio. Requires the files species_list.csv, supertree.tre, TraitsMeans.csv, and my_ultra.tre.
  • PCA.R: Creates traditional trait PCAs and weighted PCAs (see methods for more information) Requires TraitsMeans.csv and weights.csv.  and creates the files responsetrait_loadings.csv and PCA.tiff, which is Figure 2 in the paper.
  • FVT.R: Conducts FVT analyses on change in height for each species and treatment (see methods for more information). Requires Height.csv  and creates output folders height_logmodel_plots and height-expomodel_plots, and it saves this data to the dataframe height_log.csv. This data was used for Figure 3 in the paper.

The following data files are included in this submission:

  • RawData.csv: .csv file containing all the traits collected for each individual plant. Contains the following columns (see method for additional information about measurements):
    • Species: Plant species name
    • ID: Plant species ID
    • Inoculation: Inoculation treatment—Mycorrhizal (M), Non-mycorrhizal (C)
    • Block: Five blocks (A,B,C,D,E) for replicates. Block A was dropped due to shading effects that made this block significantly different than the other 4.
    • Final_Leaves: Number of leaves at final harvest; counted from leaf scans
    • Final_Height: Height of stem (or longest leaf) at final harvest (cm)
    • Leaf_area: Leaf area at final harvest (cm^2); measured by scanning leaves before drying and calculating area using ImageJ
    • LFW: Biomass (g) of leaf tissue at harvest
    • SFW: Biomass (g) of stem tissue at harvest; NA for species without stems
    • RFW: Biomass (g) of root tissue at harvest
    • Subsample_RW: Fresh root weight taken from plants for root staining check; NA for species without subsample
    • RV: Root volume (mL^3) measured using water displacement method
    • Dead_Leaves: Total number of dead leaves by harvest; tracked throughout growth period and check using leaf scars and counts of dead leaves in pot
    • Sequencing_subset: Fresh root weight taken from plants for potential sequencing; NA for species without subsample
    • LDW: Dried Biomass (g) of leaf tissue
    • SDW: Dried Biomass (g) of shoot tissue
    • RDW: Dried Biomass (g) of root tissue
    • RootLength: Root length calculated using WinRhizo root scans (m)
    • RootDiam: Average root diameter calculated using WinRhizo root scans (mm)
    • RDW_adj: Root dry weight, after including Subsample_RW and Sequencing_subset fresh weight
  • species_list.csv: Contains the species names that were used in this analysis and the synonymic names needed for the tree data files
    • Species: Species names in experiment
    • New.names: Species names that are in tree. If the original species name was not in the tree, this value will be the closest related species available.
  • supertree.tre: The super tree file obtained from Smith & Brown, 2018
  • my_ultra.tre: dichotomous tree file containing all species used in the analysis

  • weights.csv: File containing the number of replicates for each trait and species, which was used to create a weighted PCA based on certainty of each species trait measurement.

    • Species: plant species ID
    • Inoculation: inoculation treatment—Mycorrhizal (M), Non-mycorrhizal (C)
    • H: height of stem or longest leaf (cm)
    • LA: total leaf area at end of experiment (cm^2)
    • AGB: total aboveground biomass at the end of experiment (g)
    • LDMC: leaf dry matter content (g dry weight/g fresh weight)
    • SLA: specific leaf area (cm^2/g)
    • LL: leaf lifespan, measured as the number of living leaves/total leaves at the end of experiment
    • BGB: total belowground biomass at the end of experiment (g)
    • RL: root length calculated using WinRhizo root scans (m)
    • RV: root volume (mL^3) measured using water displacement method
    • RD: average root diameter calculated using WinRhizo root scans (mm)
    • RDMC: root dry matter content (g dry weight/g fresh weight)
    • SRL: specific root length (m/g)
    • RTD: root tissue density (g/mL^3)
    • RSR: root:shoot ratio
  • Height.csv: .csv file containing height measurements over time for all individuals used in this analysis

    • Species: species ID
    • Inoculation: inoculation treatment—Mycorrhizal (M), Non-mycorrhizal (C)
    • Block: species block (2-5)
    • ID: plant identifier (species+inoculation+block)
    • Height: height of stem or longest leaf (cm)
    • Days_Since_Transplant: number of days since transplant (where transplant=0)

Methods

Source material and planting

To understand the effects of mycorrhizal fungi on plant traits across the plant kingdom, we inoculated 26 prairie plant species (Table 1) with a mix of AM fungi containing Claroideoglomus claroideum (= Glomus claroideum), Funneliformus mosseae (= Glomus mosseae), Cetraspora pellucida, Claroideoglomus lamellosum, Acaulospora spinosa, Racocetra fulgida and Entrophospora infrequens (MycoBloom LLC, Lawrence, Kansas, USA). The plant species we chose for this experiment represent most of the functional and phylogenetic diversity present in the Midwest prairie ecosystem. The mycorrhizal inoculum was initially cultured from Midwest prairie soils and contains species that provide diverse benefits to hosts (Koziol and Bever, 2016). Prior to inoculation, we performed spore extractions on the inoculum to confirm spore counts and viability.

For each focal species, we scarified and cold stratified seeds for a month,then moved these seeds to the greenhouse for an additional month so that plants could germinate and grow robust enough to withstand transplantation into test pots. We filled each test pot with 1:1 sterilized soil/sand medium and maintained soil moisture at consistent levels with drip irrigation. Right before transplanting seedlings, we added mycorrhizal inoculum to each inoculated treatment pot. We chose the healthiest seedlings from each species and randomly planted two seedlings in each test pot to account for potential mortality due to transplantation and after two weeks of growth, the smallest of the two seedlings was thinned. We used two different sets of plants to measure root and leaf traits. To collect root trait data, we grew plants for 4 weeks and harvested the plants before the root systems were too large to accurately measure root traits. Plants that were used for measuring leaf traits were grown for 8 weeks so that we could track differentiation in the growth over time for each species in control and inoculated soils. Data from both sets of plants were used to calculate trait responses to mycorrhizal fungi and for the principal component analysis. For each treatment group of each species, we had four replicate pots, for a total of 16 plants per species.

Measuring plant functional traits

We measured fourteen focal functional traits in this experiment to represent five types of plant tissue response to mycorrhizal fungi: 1) change in quantity of aboveground tissues, 2) change in quality of aboveground tissues, 3) change in quantity of belowground tissues, 4) change in quality of belowground tissues, and 5) change in proportion of biomass allocation to plant tissues. We measured dried aboveground biomass (AGB), height or longest leaf in species without stems (H), and leaf area (LA) to assess aboveground tissue quantity responses to mycorrhizal fungi. To assess aboveground tissue quality responses to mycorrhizal fungi, we measured specific leaf area (SLA; ​​leaf area/leaf dry mass), leaf dry matter content (LDMC; dry leaf mass/fresh leaf mass), and leaf lifespan (LL; calculated as the proportion of living leaves at final time point to total number of leaves throughout growing period). We used SLA and LDMC as predictors of species’ placement on the leaf economic spectrum (Wright et al. 2004). We used both SLA and LDMC because, while SLA was initially described as a key element of this spectrum, the trait varies across different nutrient and light environments. Leaf dry matter content depends less on environmental conditions, so it has also been proposed as a better predictor variable for this spectrum (Hodgson et al. 2011). By measuring LL, we were able to capture the degree of photosynthetic organ turnover, which is also associated with the leaf economic spectrum (Edwards et al. 2014). To assess belowground tissue quantity response, we measured belowground dried biomass (BGB), root length (RL), root system volume (RV), and average root diameter (RD). These four metrics capture the biomass allocation belowground as well as the size of individual roots. We measured specific root length (SRL; total root length/root dried mass), root tissue density (RTD; root dry mass/root volume), and root dry matter content (RDMC; dry root mass/fresh root mass) to assess species belowground tissue quality responses. The root traits RTD and SRL represent the tradeoffs on two axes of the proposed root economic spectrum (Bergmann et al. 2020). We used RDMC as a directly analogous trait to aboveground tissue quality trait LDMC. Finally, we used RSR to assess the relative investment to roots and shoots in each plant (Qi et al. 2019). Throughout the growing period, we also measured plant height measurements (or longest leaf for plants without clearly defined stems) to track plant aboveground growth over time.

To measure changes in root traits, we grew inoculated and non-inoculated seedlings with a 1:1 sand-soil mix in 164 mL pots (cone-tainers, Stuewe and Sons, Tangent, Oregon, USA) for 4 weeks. The roots were placed on a 31×21cm water-filled tray and light scanned using an Expression 10000XL Pro scanner (Seiko Epson Corporation, Nagano, Japan). Root length and average root diameter was measured using WinRhizo software 2019a (Regent Instruments, Québec City, Quebec). We measured root volume by measuring water displacement in a graduated cylinder, which is more accurate for roots with heterogeneous diameter classes than measurements with WinRhizo (Rose, 2017). We oven dried and weighed the root and shoot tissues of these plants to calculate root to shoot ratio (RSR). We used the dried root biomass, total root length, and volume of the roots to calculate specific root length (SRL) and root tissue density (RTD). To assess aboveground traits responses to mycorrhizal fungi, inoculated and non-inoculated seedlings were grown for 8 weeks in 656 mL pots (Deepots, Stuewe and Sons, Tangent, Oregon, USA). We weighed and scanned all leaves during harvest and used ImageJ (v. 1.53) to calculate the total leaf surface area for each plant. The aboveground biomass of the plants was oven-dried and weighed to measure dried biomass. We collected a subsample of roots from inoculated and non-inoculated plants and stained them to test the efficacy of the mycorrhizal inoculation treatments. 

Data analysis

We extracted phylogenetic relationships from a larger, published phylogeny of seed plants (Smith & Brown, 2018). This phylogeny was constructed using both genetic data from GenBank and phylogenetic data from the Open Tree of Life, in order to create an inclusive, dated phylogeny of seed plants. Some species in our data set were not represented in this phylogeny, so we chose a closely related species to represent these species instead. We used the multi2di function in the ape package in R (v. 5.7.1) to convert all polytomies in this tree to dichotomies (Paradis and Schliep, 2019).

In order to understand how plant functional traits are affected by mycorrhizal fungi, we assessed the correlations between the traits for inoculated and non-inoculated plants separately. Mycorrhizal and non-mycorrhizal traits were log or square root transformed when necessary to reduce skew. To test whether our chosen functional traits exhibited phylogenetic signal across plant species, we used the phylosig function in the R package phytools (v. 1.5.1) to assess Pagel’s λ and Blomberg’s K (Revell, 2012). Because most of the traits exhibited no phylogenetic signal (Appendix S1), we chose to not use phylogenetic comparative methods in subsequent analyses. 

To test the effect of mycorrhizal fungi on these plant traits, we created a separate set of variables that were calculated as the natural logarithm of a species trait mean when grown with mycorrhizal fungi divided by the mean when grown in sterile conditions. With these metrics, positive values indicate that the trait mean increases when the species is grown with mycorrhizal fungi, compared to sterile conditions. One goal of this study was to determine whether there were coordinated responses to mycorrhizal fungi, so we ran a standard PCA with the fourteen response traits. The PCA was based on a variance-covariance matrix, because the traits for this analysis were already transformed to equivalent scales (log-response ratios), meaning that plant functional traits that have larger degrees of change between mycorrhizal and non-mycorrhizal treatments are more strongly weighted in the PCA. Due to high plant mortality, primarily in the control treatments, there were unequal sample sizes across the fourteen traits measured. To account for this, we multiplied each value in the response trait dataframe by the square root of the smallest sample size used to calculate each trait. This method preserves the direction of response but allows for data points with more confidence to be weighted stronger in the analyses. Ultimately, the weighted PCA and unweighted PCA yielded similar results, so we present the results of the weighted PCA.

We used either plant height or longest leaf in plants without stems to create function-value trait (FVT) models. In short, this method used time series data on plant growth to plot an average logistic growth curve for each species, and we calculated the response ratio of the average curve parameters to describe changes in growth over time. We dropped all individuals that did not fit logistic growth models from further analyses. We were able to fit logistic growth curves to 23 species when plants were grown with mycorrhizal fungi, compared to 13 species when grown in sterile conditions. For comparison of growth parameters, we could only include species that had logistic growth in both the inoculated and control groups. The species that were not included in this analysis were generally dropped because the growth of the control plants did not fit a logistic model due to lack of growth.We used the remaining species to calculate how response to mycorrhizal fungi changed over time by calculating maximum growth rate, time until growth curve inflection, and asymptote. Changes in growth rate indicate that mycorrhizal fungi change the rate at which the plant is able to create aboveground tissue. Decreases in inflection point when plants are colonized by mycorrhizal fungi indicate that plants reach their maximum growth rate earlier than when they were colonized by mycorrhizal fungi, whereas increases in inflection point indicate that plants are able to grow exponentially for a longer length of time. Changes in asymptote are directly analogous to changes in final size.

Funding

National Science Foundation, Award: 2037786, Graduate Research Fellowship Program

National Science Foundation, Award: 2150197, Division of Biological Infrastructure

National Science Foundation, Award: 1927696, Division of Environmental Biology

Miami University, Undergraduate Summer Scholars