Data from: Mycorrhizal identity of neighboring trees shapes seedling survival and plant-soil feedbacks through trait and light interactions
Data files
Feb 11, 2026 version files 432.78 KB
-
GH_Data.csv
426.26 KB
-
README.md
6.53 KB
Feb 11, 2026 version files 432.78 KB
-
GH_Data.csv
426.26 KB
-
README.md
6.53 KB
Abstract
Plant-soil feedbacks (PSF) play a central role in determining forest community dynamics, with trees associated with arbuscular mycorrhizal fungi (AMF) often experiencing negative PSFs, while those associated with ectomycorrhizal fungi (EMF) experiencing positive PSFs. PSFs are driven by the difference between seedling performance in both con- and hetero- specific soil, and thus the mycorrhizal match or mismatch of the heterospecific soil to the focal seedling could drive PSFs. Furthermore, functional traits and light environments could influence seedling responses. To understand how mycorrhizal type and seedling functional traits influence PSFs, we conducted a greenhouse experiment with five temperate tree species (three AM, two EM) in soils from six adult tree species (three AM, three EM). Seedlings were grown for 12 weeks under three light levels. Mycorrhizal colonization and defense/recovery traits (phenolics, lignin, nonstructural carbohydrates) were assessed at three weeks, and survival was measured over 12 weeks. AM seedlings generally experienced negative PSFs, with lower survival in AM-cultured soils, while EM seedlings had positive PSFs and higher survival in EM soils. However, negative PSFs in AM trees and positive PSFs in EM trees occurred only when the heterospecific tree culturing the soil had a mismatched mycorrhizal type relative to the focal seedling. PSFs occurred under all light levels, but their magnitude was strongest at low light, where AM seedlings showed the most negative PSFs and EM seedlings the most positive. Mycorrhizal colonization and functional trait values increased under high light and were often elevated in conspecific soils (with the exception of non-structural carbohydrates). These enhancements in traits tended to neutralize PSFs in both AM and EM seedlings. Mycorrhizal type match/mismatches between adults and seedlings influenced seedling survivorship and PSFs, likely shaping seedling recruitment patterns, and long-term forest composition, diversity, and resilience.
Dataset DOI: 10.5061/dryad.wwpzgmt0c
Description of the data and file structure
Experimental Design and Data Collection
This study used a controlled greenhouse experiment to test how plant–soil feedbacks (PSFs) influence seedling survival as a function of mycorrhizal identity, soil source, functional trait variation, and light availability. The experiment focused on how mismatches between the mycorrhizal type of focal seedlings and neighboring adult trees alter survival outcomes through soil biotic legacies and trait-mediated mechanisms.
Five temperate tree species differing in mycorrhizal association were used: Acer rubrum, Acer saccharum, Prunus serotina (arbuscular mycorrhizal; AMF-associated), and Quercus alba and Quercus rubra (ectomycorrhizal; EMF-associated). Soils were collected beneath mapped adult trees in a mixed hardwood forest and represented both conspecific and heterospecific soil sources spanning AMF- and EMF-associated species. A subset of conspecific soils was gamma-sterilized to remove soil biota without substantially altering soil chemistry or physical structure.
Seedlings were grown under contrasting light environments to test interactions between soil source and aboveground resource availability. Survival was monitored through time and analyzed using time-to-event (survival) models.
Greenhouse Experiment
The greenhouse experiment was conducted at Michigan State University’s Tree Research Center (East Lansing, Michigan, USA; 42.7°N, 84.5°W). Three light environments representing understory to gap-like conditions were created using shade cloth: low (~2% full sunlight), medium (~15% full sunlight), and high (~30% full sunlight).
Seedlings were planted into pots containing a 1:1 mixture of field-collected soil and potting substrate. Soil sources were derived from individual adult trees of known species identity. Pots were arranged across multiple greenhouse benches to account for spatial heterogeneity, with bench included as a random effect in analyses.
Seedling survival was monitored repeatedly through time. Time-to-event data were recorded, with mortality events coded explicitly and surviving individuals censored at the end of the experiment. In addition to survival outcomes, seedling functional traits associated with carbon storage and defense (non-structural carbohydrates, lignin, and phenolics) and mycorrhizal colonization metrics were measured for a sub-sample of seedlings at three weeks after planting to evaluate mechanistic pathways linking soil source to survival.
Files and variables
File: GH_Data.csv
Description: Greenhouse seedling survival experiment
Variables
-
Number: Unique seedling identifier
-
Species: Seedling species
- ACRA = Acer rubrum
- ACSA = Acer saccharum
- PRSE = Prunus serotina
- QUAL = Quercus alba
- QURU = Quercus rubra
-
SpMyco: Seedling mycorrhizal association
- AMF = arbuscular mycorrhizal
- EMF = ectomycorrhizal
-
Light: Light treatment
- Low (~2% full sunlight)
- Med (~15% full sunlight)
- High (~30% full sunlight)
-
SoilSt: Soil source type
- Nonsterile
- Sterile
-
SoilCon: Indicator of conspecific soil (binary or categorical, as used in analyses)
- Conspecific – live soil cultured by same species of adult tree as seedling
- Heterospecific – live soil cultured by different species of adult tree as seedling
- Sterile – gamma irradiated soil cultured by same species of adult tree as seedling
-
SoilSp: Species identity of soil-source adult
- ACRA = Acer rubrum
- ACSA = Acer saccharum
- PRSE = Prunus serotina
- POGR = Populus grandidentata
- QUAL = Quercus alba
- QURU = Quercus rubra
-
SoilMyco: Mycorrhizal association of soil-source adult
- AMF = arbuscular mycorrhizal fungi
- EMF = ectomycorrhizal fungi
- Sterile = gamma irradiated soil
-
SoilSpec: Soil-source category
- Con = live soil cultured by same species of adult tree as seedling
- ConSter = gamma irradiated soil cultured by same species of adult tree as seedling
- HetAMF = live soil cultured from an arbuscular mycorrhizal (AMF)-associated adult tree of a different species than the seedling
- HetEMF = live soil cultured from an ectomycorrhizal (AMF)-associated adult tree of a different species than the seedling
-
Adult: Unique identifier for adult tree soil source
-
Bench: Greenhouse bench location
-
Time: Time to event or censoring (days)
-
Event: Survival status
- 1 = mortality event
- 0 = censored (alive at end of monitoring period)
-
AMF: Arbuscular mycorrhizal colonization (%)
-
EMF: Ectomycorrhizal colonization (%)
-
NSC: Non-structural carbohydrate concentration (mg g⁻¹ dry mass)
-
Lignin: Lignin concentration (% dry mass)
-
Phenolics: Total phenolic concentration (mg g⁻¹ dry mass; gallic acid equivalents)
-
AMF_Imp: Imputed AMF colonization values (where applicable)
-
EMF_Imp: Imputed EMF colonization values (where applicable)
-
NSC_Imp: Imputed non-structural carbohydrate values
-
LIG_Imp: Imputed lignin values
-
PHN_Imp: Imputed phenolic values
Cells with no available data are coded as blanks and NA.
Code/software
All analyses were conducted within a Bayesian framework to evaluate how plant–soil feedbacks (PSFs) vary with mycorrhizal matching, soil source, and light availability. Seedling survival was analyzed using species-specific Cox proportional hazards models formulated as individual-based counting processes, with time-to-event data censored at harvest. Models included soil source and light availability as fixed effects and greenhouse bench as a random effect, and were fit in OpenBUGS (v1.4) using Markov chain Monte Carlo methods. Predicted survival probabilities were derived from posterior distributions of the hazard and risk functions, fully propagating parameter uncertainty, and PSFs were quantified as differences in predicted survival between conspecific and heterospecific soil sources within each light level. Trait expression and mycorrhizal colonization were analyzed using linear mixed models and fixed-effects ANOVA in IBM SPSS Statistics (v30), with soil source and light as fixed effects and bench as a random effect where applicable.
