Agricultural land‐use history and restoration impact soil microbial biodiversity

Abstract

Human land uses, such as agriculture, can leave long‐lasting legacies as ecosystems recover. As a consequence, active restoration may be necessary to overcome land‐use legacies; however, few studies have evaluated the joint effects of agricultural history and restoration on ecological communities. Those that have studied this joint effect have largely focused on plants and ignored other communities, such as soil microbes.
We conducted a large‐scale experiment to understand how agricultural history and restoration tree thinning affect soil bacterial and fungal communities within longleaf pine savannas of the southern United States. This experiment contained 64 pairs of remnant (no history of tillage agriculture) and post‐agricultural (reforested following abandonment from tillage agriculture >60 years prior) longleaf pine savanna plots. Plots were each 1 ha and arranged into 27 blocks to minimize land‐use decision‐making biases. We experimentally restored half of the remnant and post‐agricultural plots by thinning trees to reinstate open‐canopy savanna conditions and collected soils from all plots five growing seasons after tree thinning. We then evaluated soil bacterial and fungal communities using metabarcoding.
Agricultural history increased bacterial diversity but decreased fungal diversity, while restoration increased both bacterial and fungal diversity. Both bacterial and fungal richness were correlated with a range of environmental variables including above‐ground variables like leaf litter and plant diversity, and below‐ground variables such as soil nutrients, pH and organic matter, many of which were also impacted by agricultural history and restoration.
Fungal and bacterial community compositions were shaped by restoration and agricultural history resulting in four distinct communities across the four treatment combinations.
Synthesis and applications. Past agricultural land use has left persistent legacies on soil microbial biodiversity, even over half a century after agricultural abandonment and after intensive restoration activities. The impacts of these changes on soil microbe biodiversity could influence native plant establishment, plant productivity and other aspects of ecosystem functioning following agricultural abandonment and during restoration.

Full text

852
|
wileyonlinelibrary.com/journal/jpe J Appl Ecol. 2020;57:852–863.© 2020 British Ecological Society
Received: 5 August 2019
|
Accepted: 19 January 2020
DOI : 10.1111/1365-2664.13591
RESEARCH ARTICLE
Agricultural land-use history and restoration impact soil
microbial biodiversity
Nash E. Turley1,2 | Lukas Bell-Dereske3 | Sarah E. Evans2,3,4 | Lars A. Brudvig1,2
1Department of Plant Biology, Michigan
State Universit y, East Lansing, MI, USA
2Program in Ecology, Evolutionary Biology,
and Behavior, Michigan State University,
East Lansing, MI, USA
3Kellogg Biological Station, Michigan State
University, East Lansing, MI, USA
4Department of Integrative Biology,
Michigan State University, East Lansing, MI,
USA
Correspondence
Nash E. Turley
Email: nashuagoats@gmail.com
Present address
Nash E. Turley, Department of Biology,
University of Central Florida, Orlando, FL,
USA
Funding information
Department of Agriculture, Forest Service,
Savannah River, Grant/Award Number:
DE-EM0003622
Handling Editor: Gao wen Yang
Abstract
1. Human land uses, such as agriculture, can leave long-lasting legacies as ecosystems
recover. As a consequence, active restoration may be necessary to overcome land-
use legacies; however, few studies have evaluated the joint effects of agricultural
history and restoration on ecological communities. Those that have studied this
joint effect have largely focused on plants and ignored other communities, such
as soil microbes.
2. We conducted a large-scale experiment to understand how agricultural history
and restoration tree thinning affect soil bacterial and fungal communities within
longleaf pine savannas of the southern United States. This experiment contained
64 pairs of remnant (no history of tillage agriculture) and post-agricultural (refor-
ested following abandonment from tillage agriculture >60 years prior) longleaf
pine savanna plots. Plots were each 1 ha and arranged into 27 blocks to mini-
mize land-use decision-making biases. We experimentally restored half of the
remnant and post-agricultural plots by thinning trees to reinstate open-canopy
savanna conditions and collected soils from all plots five growing seasons after
tree thinning. We then evaluated soil bacterial and fungal communities using
metabarcoding.
3. Agricultural history increased bacterial diversity but decreased fungal diversity,
while restoration increased both bacterial and fungal diversity. Both bacterial and
fungal richness were correlated with a range of environmental variables includ-
ing above-ground variables like leaf litter and plant diversity, and below-ground
variables such as soil nutrients, pH and organic matter, many of which were also
impacted by agricultural history and restoration.
4. Fungal and bacterial community compositions were shaped by restoration and
agricultural history resulting in four distinct communities across the four treatment
combinations.
5. Synthesis and applications. Past agricultural land use has left persistent legacies
on soil microbial biodiversity, even over half a century after agricultural abandon-
ment and after intensive restoration activities. The impacts of these changes on
soil microbe biodiversity could influence native plant establishment, plant produc-
tivity and other aspects of ecosystem functioning following agricultural abandon-
ment and during restoration.

|
853
Journal of Applied Ecolog
y
TURLE Y ET aL.
1 | INTRODUCTION
The conversion of natural ecosystems for human land uses is a lead-
ing threat to biodiversity (Foley et al., 2005; Newbold et al., 2015)
and can leave long-lasting legacies on ecosystems (Flinn & Vellend,
2005; Foster et al., 20 03). For example, former agricultural lands
can support altered soils, plant communities and other proper ties
for decades to millennia following farm abandonment, relative to
‘remnant’ ecosystems with no histor y of agricultural use (Bellemare,
Motzkin, Foster, & Forest, 2002; Dupouey, Dambrine, Laffite, &
Moares, 2002; Flinn & Marks, 2007). As a consequence, active res-
toration may be necessary to overcome land-use legacies in many
ecosystems (Perring et al., 2015, 2016; Suding, 2011). Yet, our un-
derstanding of land-use legacies and the role of active restoration
for mitigating legacy effects remains unresolved for several reasons.
First, incomplete understanding results from taxonomic biases
in studies of land-use legacies and restoration. Plants have been a
strong focus in both agricultural legacy and restoration research
(Brudvig, 2011; Flinn & Vellend, 2005; Hermy & Verheyen, 2007).
Yet, taxa respond differently to legacies and restoration activities
(e.g. Jones et al., 2018) and many taxa remain poorly investigated.
Here, we focus on responses of soil microbial communities, which
can be strongly affected by land-use legacies and restoration prac-
tices (e.g. Barber, Chantos-Davidson, Amel Peralta, Sherwood, &
Swingley, 2017; Freschet, Östlund, Kichenin, & Wardle, 2014; Hui
et al., 2018; Jangind et al., 2011; Ma, De Frenne, Boon, et al., 2019;
Xue, Carrillo, Pino, Minasny, & McBratney, 2018) but are poorly
studied taxa in these fields (e.g. Brudvig, 2011).
Understanding soil microbial responses to agricultural legacies
and restoration is especially important because these groups can
have major imp acts on plant dive rsity and productivity, as well as res-
toration success (van der Bij et al., 2018; van der Heijden, Bardgett,
& Straalen, 2008; Wubs, Putten, Bosch, & Bezemer, 2016). For ex-
ample, because of host preferences and their roles in mutualistic
and antagonistic relationships, soil microbes can limit plant species
distributions and alter plant community interactions (Kardol, Martijn
Bezemer, & Putten, 2006; Wubs et al., 2016). In turn, inoculation of
former agricultural fields with mycorrhizal fungi or whole soils from
remnant ecosystems can affect plant establishment and community
assembly dynamics during restoration (Koziol et al., 2018; Wubs
et al., 2016). Thus, how agricultural legacies and restoration affect
soil microbes may have broad-reaching implications for ecosystem
recovery.
Second, studies of land-use legacies face study design challenges
(De Palma et a l., 2018), including bias es introduced through decisions
made in the past about where and how land was used by humans.
For example, temperate forests on level ground, near roads and with
higher pH soils are more likely to be converted to agricultural fields
(Flinn, Velle nd, & Marks, 20 05). In turn, fiel ds on steep slope s, located
far from roads and with lower pH soils are more likely to be aban-
doned from agriculture (Flinn et al., 2005). As a consequence, under-
lying site properties might be mistakenly interpreted as agricultural
legacy effects, when in fact they are simply consequences of the
land-use decision-making process. Controlling for these land-use
biases is particularly important for resolving how soil microbial
communities respond to land-use legacies, given soil microbes' re-
sponsiveness to soil conditions (Fierer & Jackson, 2006; Lauber,
Strickland, Bradford, & Fierer, 2008; Ma, De Frenne, Vanhellemont,
et al., 2019; Ma et al., 2018; Xue et al., 2018). Here, we control for
land-use biases through a study design where post-agricultural plots
and remnant plots with no known history of agriculture are paired in
space, resulting in no bias in underlying soil types (Brudvig, Grman,
Habeck, Orrock, & Ledvina, 2013).
Third, studies are needed to explicitly consider how restoration
affects land-use legacies and, in turn, how land-use legacies affect
restoration outcomes. Systems with a history of intensive human
land use, such as agriculture, are a common focus of restoration ef-
forts, and yet, whether and how restoration can overcome the lega-
cies of past land uses remains unclear (Jones et al., 2018; Meli et al.,
2017). Moreover, because land-use history can affect numerous
system attributes, legacies alter the template onto which restoration
acts. As a consequence, restoration outcomes may differ—perhaps
substantially—for locations with differing land-use histories (Brudvig
& Damschen, 2011; Turley & Brudvig, 2016). In other instances,
however, restoration may have clear effects that are broadly similar
to areas with and without a particular history (e.g. Breland, Turley,
Gibbs, Isaacs, & Brudvig, 2018). What is needed are controlled, repli-
cated experiments to draw strong inferences about how restoration
and land-use legacies interact. Yet experiments manipulating resto-
ration treatments across areas differing in land-use history are rare.
Ideally, such investigations would be coupled with measurements of
key environmental variables hypothesized to affect the taxa of inter-
est, including those affected by agricultural history and restoration.
For example, within our focal system, bee responses to restoration
may be mediated by the increase in flower cover resulting from res-
toration (Breland et al., 2018). Therefore, determining mechanisms
of legacy and restoration effects likely requires coupled measure-
ments of key environmental variables along with the focal taxa.
We overcame these limitations through a replicated restoration
experiment in longleaf pine savannas. Our experiment included a
factorial manipulation of restoration (overstorey tree thinning) and
agricultural history, whereby plots with and without a history of till-
age agriculture received restoration or were left as unrestored con-
trols. We arranged plots into blocks to control for land-use biases
and considered how agricultural history, restoration thinning and
their interaction affect soil bacteria and fungi.
KEY WORDS
agricultural history, bacteria, community ecology, fungi, land-use legacy, metabarcoding,
restoration, soil microbe biodiversity

854
|
Journal of Applied Ecology
TURLE Y ET aL.
Previous work within our experiment has shown how agricultural
history and restoration affect abiotic and biotic conditions in ways
that we expect to influence soil microbial communities and diver-
sity. In particular, compared to remnants, post-agricultural savannas
support soils that are more compacted, support elevated phospho-
rus and reduced organic matter content and water holding capac-
ity, as well as lower tree canopy cover and altered plant community
composition (but comparable plant species richness; Brudvig et al.,
2013). The restoration thinning treatment decreased tree canopy
cover and litter accumulation; increased near-ground temperatures
and sunlight reaching ground level; increased plant species richness;
and altered plant community composition (Hahn & Orrock, 2015;
Stuhler & Orrock, 2016; Turley & Brudvig, 2016). Based on these
past findings, we suspected that soil microbial diversity and commu-
nity composition will also be affected by both land-use history and
restoration thinning.
We asked the following questions related to both soil bacteria
and soil fungi from our field experiment:
1. What are the effects of agricultural land-use history and res-
toration on soil microbe diversity and composition?
2. Does the effect of restoration on diversity and composition de-
pend on agricultural land-use history (i.e. do agricultural history
and restoration interact)?
3. Do environmental variables correlate with microbial biodiversity?
4. Do correlations between microbe diversity metrics and environ-
mental variables help explain the impacts of restoration and land-
use history on soil microbe biodiversity?
2 | MATERIALS AND METHODS
2.1 | Study location and experimental design
Our research took place at the Savannah River Site (SRS), an
~80,000 ha National Environmental Research Park located on the
upper coastal plain in South Carolina (33.20°N, 81.40°W). This area
historically supported fire-maintained longleaf pine savanna in the
sandy uplands (Kilgo & Blake, 20 05)—an ecosystem characterized
by sparse canopies dominated by longleaf pine trees Pinus palustris
and a dense understorey plant layer of graminoids, forbs and shrubs
(Noss et al., 2015). By the mid-20th centur y, most of the SRS uplands
had been converted to tillage agriculture, primarily for cotton and
corn (Kilgo & Blake, 2005). In 1951 the US government obtained SRS
and began converting agricultural fields to longleaf, loblolly Pinus
taeda and slash pine Pinus elliottii plantations (Kilgo & Blake, 2005).
Following acquisition (and likely for decades prior to this), fire was
excluded from ecosystems within SRS, until initiation of prescribed
burning in the early 21st century (Kilgo & Blake, 2005).
At SRS, we conducted a factorial experimental manipulation of
agricultural history and restoration tree thinning, across 126 1-ha
plots arranged into 27 blocks (Figure 1). Each block was focused
around a fragment of remnant longleaf pine savanna, with no known
history of tillage agriculture, adjacent to a former agricultural field
supporting closed-canopy pine (longleaf where possible) plantation
at the initiation of the study (Brudvig et al., 2013). We determined
land-use histories for each plot using historical aerial photos taken
in 1951, at the time of SRS's creation (Brudvig et al., 2013). Remnant
and post-agricultural plots within blocks supported similar soil types
and topographies (Brudvig et al., 2013), suggesting that the blocked
experimental design adequately controlled for non-random land-use
decision-making.
In 2011, prior to the start of the growing season, we applied a
tree thinning treatment to restore open-canopy, savanna structure
to half of the remnant and post-agricultural plots (Turley & Brudvig,
2016). This reduced tree densities from an average of 650 trees/ha
to 10 trees/ha. All plots have subsequently been managed with one
or more prescribed fires. The frequency of prescribed surface fire
did vary among the 26 blocks since the initiation of the experiment;
however, all plots, and thus all four treatment combinations, within
a block were always burned together. Although fire could be an im-
portant factor shaping soil microbes within longleaf pine savannas
(Semenova-Nelsen, Platt, Patterson, Huffman, & Sikes, 2019) look-
ing at this is beyond the scope of this study.
FIGURE 1 Diagram showing the
experimental sites and soil sampling
locations, within the Savannah River
Site in South Carolina. Each of 27 sites
has 1-ha experimental plots in remnant
and post-agricultural areas. Half of the
1-ha plots in each land-use type had
restoration tree thinning in 2011 to
restore open-canopy savanna conditions.
Soil samples were collected across all 1-ha
plots in 2015
RemnantPost-
agricultural
Soil sample location
Vegetation transect
Thinned
Thinned Control
20 km
100 m
1 m
Single soil probe
South Carolin
a
Savanna River Site
1 of 27 experimental sites
Control

|
855
Journal of Applied Ecolog
y
TURLE Y ET aL.
2.2 | Soil sampling and processing
In Fall 2015 we collected ~1.2 L of soil from each of the 126 1-ha
plots. Each soil sample was an aggregate of 30 1.6 cm wide by
20 cm deep soil probes collected along t wo 50 m transects through
the middle of each plot (Figure 1). The soil sampling transects ran
on both sides of our already-present vegetation sampling transects
(Turley & Brudvig, 2016; Figure 1). Before each probe the leaf lit-
ter, duff and sticks were brushed aside. To minimize contamina-
tion we used one soil probe for all remnant sites and another for
all post-agricultural sites and between each plot we rinsed the in-
side and outside of the probe with a 10% bleach solution and then
water. Aggregate soil samples were mixed thoroughly and split up
for different purposes. About 50 ml was stored in a −20°C freezer
for microbial analysis, and two other subsamples were used for en-
vironmental sampling.
For microbial analysis, we extracted soil DNA using MoBio
PowerSoil Extraction Kit following the manufacturer's instructions.
We submitte d DNA to the Michigan St ate University Core Genomics
Facility for Illumina sequence librar y construction. Following their
standard protocols, bacterial 16S V4 (515f/806r) and ITS (ITS-F/
ITS2) Illumina compatible libraries were prepared using primers
containing both the target sequences and the dual indexed Illumina
compatible adapters. The 16S and ITS1 amplicon pools were se-
quenced independently in a 2 × 250bp paired end format using in-
dependent v2 500 cycle MiSeq reagent cartridges.
The first of the soil subsamples was analysed by Brookside
Laboratories Inc. for soil texture (percent sand, clay and silt), pH,
organic matter, and nutrients and minerals. On the second subsam-
ple we measured soil water holding capacity (proportionate differ-
ence between saturated wet and oven dry weight) and gravimetric
soil moisture using the same methods as Brudvig and Damschen
(2011). Soil pH, water holding capacity, organic matter and several
soil nutrients all decreased with agricultural history while soil phos-
phorus was strongly increased (see Table S1).
2.3 | Environmental data collection
We measured a set of environmental variables within each experi-
mental plot at 10 m intervals along the 100 m vegetation transects
(Figure 1) during the 2015 growing season. In 1 × 1 m plots we visu-
ally estimated the percent cover of leaf litter, down woody debris,
bare ground and understorey vegetation. At each of these plots we
also measured the depth of leaf litter and canopy cover of oversto-
rey trees using a spherical densiometer. In 1 × 1 m and 10 × 10 m
plots we recorded all plant species and calculated plant species
richness. For all these environmental variables we averaged the 10
measurements across each transect to get one value per 1-ha plot.
Restoration thinning resulted in strong declines in leaf litter and
canopy cover and large increases in vegetation cover and under-
storey plant richness (Table S1). Units and methods for measuring
all of our environmental variables are available in Table S6.
2.4 | Bioinformatics
We processed and clustered bacterial and fungal reads into
operational taxonomic units (OTUs). Reads from the bacterial
community were chimera checked, quality filtered and merged
using Trimmomatic and Pandaseq (Bolger, Lohse, & Usadel,
2014; Masella, Bartram, Truszkowski, Brown, & Neufeld, 2012).
Processed reads were clustered into OTUs at 97% identity
level using UCLUST6.1 with the default settings (Edgar, 2010).
Singletons were removed and contigs were screened using QIIME
1.9.1 (Caporaso et al., 2010) with the default parameters. OTUs
classified to chloroplast, mitochondria or with less than four
reads across all samples were filtered out to avoid over splitting
(Thiéry, Moora, Vasar, Zobel, & Öpik, 2012) and sequencing errors
(Dickie, 2010). The resulting community was composed of 90,103
OTUs and 1,650,420 reads. Fungal reads were quality filtered
and merged using the USEARCHv10 pipeline (http://drive5.com/
usear ch/; Edgar, 2010, 2013). Merged sequences were quality
filtered to an expected error threshold of 1.0 fastq_filter (Edgar
& Flyvbjerg, 2015) and primer sequences bases were removed.
The combined reads were clustered into OTUs at 97% identity
level then reference-based chimera checked (Edgar, 2016) and
classified against the UNITE 7.1 ITS1 chimera and reference da-
tabases respectively (Kõljalg et al., 2013). All non-fungal OTUs
and those with less than four reads were filtered from the com-
munity matrix. The resulting fungal community had 10,285 OTUs
and 584,113 reads.
2.5 | Statistical analysis
We conducted all analyses in R version 3.5.1. We first removed
two samples with extremely low reads: a bacteria sample with
471 reads and a fungal sample with 78 reads (compared to means
of ~69,000 and 5,000 respectively). For measuring diversity we
rarified the community datasets following Weiss et al. (2017) using
the ‘rrarefy’ function (Oksanen et al., 2010). We set the minimum
value in the rarefaction to the lowest observed read number in a
sample. With those community datasets we calculated richness,
evenness and inverse Simpson's D. Our evenness metric was in-
verse Simpson's diversity divided by species richness. We focus
primarily on inverse Simpson's D as our measure of biodiversity
as this is recommended for microbial datasets (Haegeman et al.,
2013). We evaluated correlations between average plot-level (1-ha)
environmental variables and diversity metrics using Pearson's
correlations.
To test the effects of agricultural history and restoration thin-
ning on biodiversity metrics we fit mixed effects models using the
‘lmer’ function (Bates, Mächler, Bolker, & Walker, 2015). We included
restoration thinning, agricultural history (both two-level factors) and
their interaction as fixed effects. Site (a 27-level categorical factor)
and land-use histor y were included as random effects. Land-use
history was nested within site to account for the pseudoreplication

856
|
Journal of Applied Ecology
TURLE Y ET aL.
inherent in the hierarchical experimental design. The model syntax
was:
We used the ‘ANOVA’ function (Fox & Weisberg, 2018) to calcu-
late p-values using Type 2 sums of squares. We used Type 2 sums of
squares because our models had non-significant interaction terms and
this allowed us to interpret the main effects while keeping the inter-
action term in the model. We determined R2 for the factors using the
‘r2beta’ function with the standardized generalized variance method
(Jaeger, 2017). For community composition analyses we transformed
the data using the ‘varianceStabilizingTransformation’ function with
the ‘local’ fit type (Love, Huber, & Anders, 2014; Weiss et al., 2017).
On the transformed datasets we created a distance matrix using Bray–
Curtis dissimilarity which was abundance weighted by read number.
We tested the effects of our factors on community composition by
fitting PERMANOVA models with the ‘adonis’ function (Oksanen et al.,
2010). We included the site factor as a ‘strata’ term. Because nesting
is not possible with the ‘adonis’ function the degrees of freedom for
these tests are inflated which could artificially reduce p-values. We vi-
sualized the effects of our treatments on community composition by
performing a constrained analysis of principal coordinates using the
‘capscale’ function with default parameters then visualizing the ordi-
nation using the ‘ordiplot’ function (Oksanen et al., 2010). We used the
‘envfit’ function (Oksanen et al., 2010) to test for correlations bet ween
environmental variables (Bray–Curtis dissimilarity matrix) and the mi-
crobe community ordinations (non-metric multidimensional scaling
with Bray–Curtis dissimilarity). To account for concerns of oversplit-
ting due to open reference OTU clustering (Edgar, 2017), we ran the
same PERMANOVA model on the bacterial Unifrac distance matrix.
Accounting for phylogeny did not change the results, so we only pres-
ent the Bray–Curtis-based results.
We explored the relationship among experimental treatments,
environmental variables and microbial diversity variables using
structural equation modelling. Because there were many, some-
times collinear, potential environmental variables to include in
the analyses (Tables S3 and S4), we simplified the data into two
composite variables using a principle components analysis (PCA).
We standardized all variables to have a mean of 0 and standard
deviation of 1, then fit SEM’s using the ‘sem’ function (Rosseel,
2012). We fit models with PC1 and PC2 as endogenous variables
between the treatments and microbe biodiversity metrics. To test
the importance of the environmental variables (PC1 and PC2) in
the models, we fit SEM’s without them and compared the R2 to the
full models with them included.
3 | RESULTS
3.1 | Question 1: Effects of agricultural history and
restoration on soil microbial biodiversity
History of agricultural land use had opposite effects on bacterial
and fungal diversity (inverse Simpson's D) and also shaped com-
munity composition. For bacteria, agricultural history increased
diversity by 53.7% (Figure 2a; Table 1) whereas for fungi, agri-
cultural history reduced diversity by 18.5% (Figure 2b; Table 1).
These results were driven primarily by changes in evenness for
bacteria and richness in fungi (Table S2). Agricultural history also
significantly affected microbial composition (Figure 3; Table 1)
which explained 2.5% of bacterial and 3.9% of fungal community
variation.
Restoration increased both bacterial and fungal diversity and
impacted community composition. Restoration increased bacterial
diversity by 13.8% (Figure 2a; Table 1) and fungal diversity by 60.1%
(Figure 2b; Table 1). These changes in diversity were driven by in-
creases in both richness and evenness (Table S2). Restoration thin-
ning also shaped bacterial and fungal communities (Figure 3; Table 1)
and this factor explained 1.2% and 2.6% of variation in communities
respectively.
3.2 | Question 2: Effects of agricultural history on
restoration effects
Overall there was little evidence that the effects of restoration
were dependent on agricultural history. There were no significant
interactions between restoration and agricultural history for bacte-
rial or fungal diversity (Table 1). There was a significant interaction
between agricultural history and restoration on fungal community
composition explaining 1% of variation.
Y
∼thinning ∗land use +
(
1
|
site∕land use∕thinning
).
FIGURE 2 Effects of agricultural land-
use history and restoration thinning on
diversity (inverse Simpsons's D) within
a longleaf pine savanna experiment in
South Carolina for (a) bacteria and
(b) fungi. Remnant plots are savannas
with no history of agriculture and post-
agricultural sites had tillage agriculture
that was abandoned over 60 years ago
and then managed as pine plantation
(a) (b)

|
857
Journal of Applied Ecolog
y
TURLE Y ET aL.
3.3 | Question 3: Correlations between
environmental variables and soil microbial
biodiversity
Many environmental variables were correlated with soil microbial di-
versity, richness and evenness (Tables S3 and S4). A PCA collapsed
this variation into two composite variables. The first axis from this
analysis (PC1) was associated mostly with below-ground variables.
Negative values were associated with % sand, soil Fe and soil P, while
positive values were associated with a wide range of soil micronutri-
ents, soil organic matter, soil water holding capacity, % silt and soil pH
(Figure 4; Table S5). PC2 was associated mostly with above-ground
variables related to canopy density. Positive values of PC2 were as-
sociated with % canopy cover, leaf litter and soil S, whereas negative
values were associated with % cover bare ground (Figure 4; Table S5).
Plant richness, % vegetation cover and % leaf litter were associated
with both a xes, with PC1 positively associated with plant richn ess and
PC2 negatively associated with plant richness (Figure 4; Table S5).
The principle components of environmental variables pre-
dicted soil microbial richness and evenness, and diversity. The
strongest correlations were between PC1 and richness (Table 2).
TABLE 1 Results of models for soil bacteria and fungal
Simpson's diversity and community composition from longleaf
pine savannas. Data are from an experiment with 126 1-ha
plots, factorially manipulating agricultural land-use history and
restoration tree thinning. Inverse Simpson's diversity results
are from mixed effects models and community results are from
multivariate PERMANOVA models
ddf F p r2
Bacteria
Inverse Simpson's D
Land use 23 148.95 <.0 01 .390
Restoration 46 16.25 <.001 .048
Land use × rest. 45 0.22 .639 .002
Community
Land use 121 3.14 <.001 .025
Restoration 121 1.48 .002 .012
Land use × rest. 121 0.98 .354 .008
Fungi
Inverse Simpson's D
Land use 25 5.44 .028 .003
Restoration 50 32.57 <.001 .079
Land use × rest. 50 2.23 .142 .021
Community
Land use 121 5.13 <.0 01 .039
Restoration 121 3.40 <.0 01 .026
Land use × rest. 121 1.36 .034 .010
Note: DDF, denominator degrees of freedom.
Values with p < .05 are bolded.
FIGURE 3 Effects of agricultural land-use history and
restoration thinning on: (a) bacteria community composition and
(b) fungal community composition, from longleaf pine savanna soils
CAP1
CAP2
210
12
210 12
CAP1
CAP2
0.3 0.1 0.1 0.3
0.4 0.2 0.0 0.2
Post-ag./control
Post-ag./thinned
Remnant/control
Remnant/thinned
Bacteria Fungi
(a) (b)
FIGURE 4 Ordination of a principle components analyses of all
environmental variables collected from each of the 126 1-ha plots.
The location of each variable along each axis indicates how strongly
associated the variable is with that axis. PC1 is strongly associated
with various below-ground variables such as nutrients, soil texture and
soil moisture. PC2 is most associated with above-ground variables like
tree canopy cover, leaf litter and bare ground. However, both axes are
strongly associated with plant richness and percent cover of vegetation
PC1
PC2
P.veg
P.litter
P.wood
P.tree.trunk
P.bare.ground
P.canopy.cover
Litter.depth
Duff.depth
richness.1 × 1
richness.10 × 10
P.clay P.silt
P.sand
pH
OM
S
P
Ca Mg
K
Na
Fe
Mn
Cu
Al
P.moisture PWHC
–1.0 –0.50.0 0.51.0 1.5
–0.5 0.00.5 1.0
TABLE 2 Pearson's correlations between soil microbe
biodiversity metrics and principle component axes of soil and
vegetation environmental parameters (see Figure 4). All variables
were measured within 126 1-ha longleaf pine savanna
Variable 1 Variable 2
Bacteria Fungi
r p r p
Richness PC1 .46 <.001 .66 <. 001
Evenness PC1 −.18 .05 .11 .21
Simpson's D PC1 −.02 .78 .37 <.0 01
Richness PC2 −.21 .02 −.14 .13
Evenness PC2 −.35 <.0 01 −.33 <.0 01
Simpson's D PC2 −.37 <.0 01 −. 31 <.0 01
Note: Values with p < .05 are bolded.

858
|
Journal of Applied Ecology
TURLE Y ET aL.
Locations with wetter, more nutrient rich and basic soils, and
with greater plant species richness supported greater soil mi-
crobial richness, and this relationship was stronger in fungi than
in bacteria (Figure 5). PC1 was somewhat negatively correlated
with bacterial evenness and not significantly correlated with di-
versity (Table 2). PC1 had no relationship with fungal evenness
and was positively correlated with fungal richness (Table 2). PC2
was negatively correlated with all measures of fungal and bac-
terial biodiversity (richness, evenness and Simpson's diversity)
with the exception of fungal richness (Table 2). Thus, plots with
greater tree canopy cover and leaf litter had reduced soil mi-
crobial diversity, whereas plots with more bare ground, under-
storey vegetation and plant richness supported greater levels of
microbial biodiversity.
Microbial community composition was also correlated with a
wide range of environmental variables (Tables S3 and S4). Bacterial
communities were correlated with most below-ground variables
such as soil pH, nutrients, texture and water holding capacity
(Table S3), but not with above-ground variables (with the exception
of one measure of plant richness). Fungal communities were also
correlate d with below-ground variables, similar to bacteria, but were
also correlated with above-ground variables such as plant richness,
leaf litter and tree canopy cover (Table S4). Overall, environmental
variables had significant correlations with community ordination
for bacteria (Mantel test, r = .21, p = .001) and fungi (Mantel test,
r = .23, p = .0 01).
3.4 | Question 4: Do environmental variables help
explain effects of treatments on microbe biodiversity
Our structural equation models (SEM’s) showed that agricultural
land-use history and restoration treatments impacted microbial di-
versity (inverse Simpson's D) and evenness mostly independently of
the environmental variables we measured, while microbial richness
was mostly predicted by environmental variables and not the treat-
ments. The SEM’s showed that agricultural history and restoration
thinning impacted both of the environmental PC axes (Figure 6) and
the direct effects of the treatments on environmental variables are
summarized in Table S1. Agricultural history was the strongest pre-
dictor of bacterial diversity, but the environmental variables were
also significant (Figure 6a). The model overall explained 57% of the
variation in bacterial diversity (Figure 6a). A SEM fit without the
environmental variables as intermediates between the treatments
and diversity still explained 53% of variation in bacterial diversity.
The fungal diversity SEM had restoration thinning as a significant
FIGURE 5 Relationship between the first principle component axis of environmental variables (see Figure 4) on (a) bacterial richness
and (b) fungal richness. Richness was calculated from a rarefied community dataset. Negative values of PC1 are associated with % sand, Fe,
P, leaf litter while positive values are associated with a wide range of soil micronutrients, soil organic matter, soil water holding capacity, %
vegetation cover and plant richness
3,000
3,500
4,000
4,500
–1 012
Environmental PC1
Richness
200
300
400
500
600
–1 012
Environmental PC1
Richness
Post-ag./control
Post-ag./thinned
Remnant/control
Remnant/thinned
(a) Bacteria (b) Fungi
r = .46, p < .001 r = .66, p < .001
FIGURE 6 Structural equation
model path diagrams showing the main
treatment effects at the top, principle
component axis of environmental
variables in the middle and inverse
Simpson's diversity at the bottom for (a)
bacteria and (b) fungi. The width of the
arrows is proportional to the magnitude
of the path coefficient. Black arrows are
positive correlations, grey arrows are
negative correlations and dashed arrows
are non-significant paths
(a) (b)

|
859
Journal of Applied Ecolog
y
TURLE Y ET aL.
predictor along with the environmental PC axes (Figure 6b) which
explained a total of 30% of the variation in diversity. This model
without the environmental variables explained 24% of variation in
fungal diversity. The models for evenness (both for bacterial and
fungal) showed similar patterns to those of diversity with the envi-
ronmental variables explaining minimal variation (<2%) in evenness
(Figure S4).
SEM explained little variation in microbial richness when environ-
mental variables were excluded. The full model for bacterial richness
explained 44% of variation in richness (Figure S3) but without environ-
mental variables explained only 4%. Similarly, for fungal richness the
full model explained 48% of variation in richness (Figure S3) while the
model without environmental variables explained only 17%.
4 | DISCUSSION
Soil bacteria and f ungi biodiversit y were both affected by agricultural
history, restoration thinning and environmental variables. Our results
point to four major conclusions: (a) agricultural history increased bac-
terial diversity while reducing fungal diversity, (b) restoration thin-
ning increased fungal and bacterial diversity, (c) agricultural history
and restoration thinning resulted in four distinct bacterial and fungal
communities across the four plot types and (d) environmental varia-
bles were important predictors of microbial diversity, mostly through
their impacts on microbial richness.
4.1 | Possible explanations for changes in
bacterial and fungal biodiversity
Agricultural land-use history increased bacterial diversity, similar
to findings from other studies (Delgado-Baquerizo et al., 2017;
Dong, Huai-Ying, De-Yong, & Huang, 2008; Hartman, Richardson,
Vilgalys, & Bruland, 2008; Jesus, Marsh, Tiedje, & Moreira, 2009;
Rodrigues et al., 2013; Upchurch et al., 2008). Soil nutrients
(Delgado-Baquerizo et al., 2017; Lauber et al., 2008) and soil pH
(Jesus et al., 2009; Rodrigues et al., 2013) may be important fac-
tors mediating land-use history effects on microbial diversity.
Similarly, we found a suite of variables that correlated with bacte-
rial diversity (Figure 6) and richness (Figure 5) that were also im-
pacted by agricultural history. In our system, post-agricultural sites
had decreased soil organic matter, micronutrients (S, Ca, Mg, Al
and K), moisture, and water holding capacity and increased soil P
(Table S1). Given collinearities among these variables (Figure 4), it
is difficult to say which of those that correlated with measures of
bacterial metrics of biodiversity (Table S3) mechanistically influ-
enced diversity. However, we did find a strong pattern that envi-
ronmental variables, especially below-ground variables, were the
most important predictors of bacterial richness, greatly increas-
ing our predictive power of the effects of treatments on richness
(Figure S3). However environmental variables explained much less
variation in diversity (Figure 6), and almost none at all for evenness
(Figure S4). This suggests that microbial evenness and richness are
responding to fundamentally different environmental gradients in
this system and illustrates the importance of considering multiple
biodiversity measures when evaluating responses to disturbance
and management.
In contrast to bacteria, fungal diversity was lower in post-
agricultural plots, although the magnitude of this response was rel-
atively small (Figure 2). Other studies have also found that agricul-
tural land use lowers fungal diversity (Ding et al., 2013; Oehl et al.,
2003; Wagg, Dudenhöffer, Widmer, & Heijden, 2018) and our anal-
yses suggest that the above-mentioned environmental variables
associated with bacteria could also be important factors shaping
fungal diversity. It is also possible that post-agricultural recovery
was limited by dispersal from remnant to post-agricultural plots
for fungi, as we see for plants (Turley, Orrock, Ledvina, & Brudvig,
2017), or that fungi are relatively slower growing than bacterial and
thus slower to recover following disturbance.
Restoration increased both bacterial and fungal diversity, al-
though the effect was stronger for fungi (Figure 2). Decreases in
canopy cover and leaf litter, along with increases in vegetation
cover and plant richness, may help explain the increased bacterial
richness and diversity in thinned plots as PC2 was a strong pre-
dictor of bacterial diversity (Figure 6a) and richness (Figure S3).
However, this was less for fungi (Figure 6b; Figure S3). Restoration
greatly increased plant species richness (Table S1; Turley & Brudvig,
2016), which may mediate the effects of restoration thinning on
soil microbial communities by increasing the number of suitable
plant hosts for host-specific microbes (Peay, Baraloto, & Fine,
2013; Prober et al., 2015), although it is also possible that microbial
diversity enhanced plant richness. Finally, restoration thinning in
savanna ecosystems can increase the variability in biota and en-
vironmental gradients (Brudvig & Asbjornsen, 2009), thereby in-
creasing the number of potential niches within a site, for microbes
of diverse life histories (Curd, Martiny, Li, & Smith, 2018). Such en-
hancement of heterogeneity may be particularly important when
restoring post-agricultural ecosystems, like in our study, given re-
ductions in heterogeneity that can persist for decades or longer
following agricultural abandonment (Flinn & Marks, 2007).
4.2 | Community composition in response to
agricultural history and restoration
Our results illustrate how agricultural legacies are long-lasting for
soil microbial communities, persisting over half a century after ag-
ricultural abandonment despite post-agricultural and remnant plots
being adjacent in our experiment. These findings add to a grow-
ing body of literature showing varying effects of land-use legacies
on soil microbes (Fichtner, Oheimb, Härdtle, Wilken, & Gutknecht,
2014; Hartman et al., 2008; Hui et al., 2018; Jangind et al., 2011;
Lauber et al., 2008; Upchurch et al., 2008), although some studies
show no impacts of land-use history on soil bacteria (Ma, De Frenne,
Boon, et al., 2019; Ma, De Frenne, Vanhellemont, et al., 2019). Our

860
|
Journal of Applied Ecology
TURLE Y ET aL.
community analyses show that both fungal and microbial communi-
ties cluster into four distinct community types (Figure 3; Table 1),
which is very similar to how plant communities have responded to
our treatments (Turley & Brudvig, 2016). This means that restoration
did not result in post-agricultural communities being more similar to
remnant communities. Similarly, Strickland et al. (2017) found that
restored forests in Mississippi had soil microbial communities dis-
tinct from agricultural fields and from nearby remnant forests. They
conclude that above-ground restoration focused on forest structure
does little to drive microbial communities towards the remnant ref-
erence state, or perhaps that these changes will happen very slowly
or be contingent on restoration of plant community composition.
Alternatively, agricultural legacies could be due to priority effects
where chance events early in community assembly results in differ-
ent community outcomes that persist even with the recovery of en-
vironmental conditions (Keiser, Strickland, Fierer, & Bradford, 2011).
4.3 | Implications for management
We found little evidence that the effects of restoration thinning
for soil microbes differed between remnant and post-agricultural
plots. This finding suggests that agricultural history and restora-
tion are independently operating on different groups of microbial
species, with some species either dispersal limited or affected by
altered environmental gradients following agricultural abandon-
ment (e.g. elevated soil phosphorus) and a second group promoted
by restoration thinning. This presents a mixed message for the
prospects of soil microbial recovery during restoration. On the one
hand, restoration can increase the diversity of soil fungi and bac-
teria in plots within either land-use history. On the other hand,
restoration does not mitigate the legacies of historical agricultural
land use. Thus, successful soil microbial restoration may require
coupling of structural habitat manipulation to reinstate appropri-
ate environmental conditions for a diverse suite of microbes with
active reintroduction of soil microbes that do not recover passively
following agricultural land use (e.g. Koziol et al., 2018; Wubs et al.,
2016). In turn, active reintroduction of soil microbes may be im-
portant for re-establishing certain plant species during restoration
(Harris, 2009; Kardol & Wardle, 2010). Evidence to date from our
experiment does not support this, however, with a suite of under-
storey herbs actually establishing better in post-agricultural plots
and performing similarly when grown in soils inoculated with soil
microbes from remnant and post-agricultural plots (Barker, Turley,
Orrock, Ledvina, & Brudvig, 2019).
Whether and how soil microbial communities recover following
human land use and ac tive restoration efforts remains an open ques-
tion (Harris, 2009) and our study adds to accumulating evidence that
restoration actions manipulating ecosystem structure and plant di-
versity (directly or indirectly) also affect soil microbial communities
(Banning et al., 2011; Barber et al., 2017; Dickens, Allen, Santiago, &
Crowley, 2015; Potthoff et al., 2006). We further illustrate the po-
tential for r estoration to ben efit soil microbes across site s supporting
different land-use histories. Given the consequences of microbial
communities for ecosystem dynamics during restoration (Kardol
& Wardle, 2010), soil microbial differences resulting from land-use
legacies and restoration actions may have broad-reaching implica-
tions for ecosystem recovery and restoration outcomes in degraded
ecosystems.
ACKNOWLEDGEMENTS
We are indebted to John Blake, Andy Horcher, Ed Olson and the
prescribed fire crew at the USDA Forest Service-Savannah River for
their assistance with creating and maintaining the Remnant Project
experiment. We thank Sabrie Breland, Joe Ledvina and John Orrock
for their help with coordinating the Remnant Project experiment,
Selina Pradhan for laboratory assistance and Will West (Evans Lab)
for assistance with bioinformatics. This work was supported by
funds provided to the Department of Agriculture, Forest Service,
Savannah River, under Interagency Agreement DE-EM0003622
with the Department of Energy, Aiken, SC.
AUTHORS' CONTRIBUTIONS
N.E.T. and L.A.B. conceived the research idea and wrote the paper;
N.E.T. collected the field samples and analysed the data; L.B.-D. and
S.E.E. conduc ted laboratory work and bioinformatics. All the authors
edited the paper.
DATA AVA ILAB ILITY STATE MEN T
All raw sequence data from this study are available through the
NCBI Sequence Read Archive under project PR JNA551504 and
SRAs SRR9609456 - SRR9609568. Data available via the Dryad
Digital Repository https://doi.org/10.5061/dryad.x3ffb g7fd (Turley,
Brudvig, Bell-Dereske, & Evans, 2020).
ORCID
Nash E. Turley https://orcid.org/0000-0001-7318-8786
REFERENCES
Banning, N. C., Gleeson, D. B., Grigg, A. H., Grant, C. D., Andersen, G. L.,
Brodie, E. L., & Murphy, D. V. (2011). Soil microbial community suc-
cessional patterns during forest ecosystem restoration. Applied and
Environmental Microbiology, 77, 6158–6164. htt ps://doi.org/10.1128/
AE M . 0 076 4 -11
Barber, N. A., Chantos-Davidson, K. M., Amel Peralta, R., Sherwood, J.
P., & Swingley, W. D. (2017). Soil microbial community composition
in tallgrass prairie restorations converge with remnants across a 27-
year chronosequence. Environmental Microbiology, 19, 3118–3131.
https://doi.or g/10.1111/1462-2920.13785
Barker, C. A., Turley, N. E., Orrock, J. L ., Ledvina, J. A., & Brudvig , L . A.
(2019). Agricultural land-use histor y does not reduce woodland un-
derstory herb establishment. Oecologia, 89, 1049–1060. https://doi.
org /10.1007/s004 42- 019-0 43 48-6
Bates, D., Mächler, M., Bolker, B., & Walker, W. (2015). Fitting linear
mixed-effects models using lme4. Journal of Statistical Sof tware,
67(1), 1–48. https://doi.org/10.18637/jss.v067.i01
Bellemare, J., Motzkin, G ., Foster, D. R., & Forest , H. (20 02). Legacies
of the agricultural past in the forested present: An assessment
of historical land-use effects on rich mesic forests. Journal of

|
861
Journal of Applied Ecolog
y
TURLE Y ET aL.
Biogeography, 29, 1401–1420. https://doi.org/10.1046/j.1365-2699.
2002.00762.x
Bolger, A. M., Lohse, M., & Usadel, B. (2014). Trimmomatic: A flexible
trimmer for Illumina sequence data. Bioinformatics, 30, 2114–2120.
https://doi.org/10.1093/bioin forma tics/btu170
Breland, S., Turley, N. E., Gibbs, J., Isaacs, R., & Brudvig, L. A. (2018).
Restoration increases bee abundance and richness but not pollination
in remnant and post-agricultural longleaf pine woodlands. Ecosphere,
9, e02435.
Brudv ig, L. A. (2011). The re storation of biodiversi ty: Where has research
been and where does it need to go? American Journal of Botany, 98,
549–558. https://doi.org/10.3732/ajb.1000285
Brudvig, L. A ., & Asbjornsen, H. (2009). The removal of woody encroach-
ment restores biophysical gradients in Midwestern oak savannas.
Journal of Applied Ecology, 46, 231–240. https://doi.org/10.1111/
j.1365-2664.2008.01590.x
Brudvig, L. A ., & Damschen, E. I. (2011). Land-use history, historical con-
nectivity, and land management interact to determine longleaf pine
woodland understory richness and composition. Ecography, 34, 25 7–
266. https://doi.org/10.1111/j.1600-0587.2010.06381.x
Brudvig, L. A., Grman, E., Habeck, C. W., Orrock, J. L., & Ledvina, J.
A. (2013). Strong legacy of agricultural land use on soils and un-
derstory plant communities in longleaf pine woodlands. Forest
Ecology and Management, 310, 944–955. https://doi.org/10.1016/j.
foreco.2013.09.053
Caporaso, J. G., Kuczynski, J., Stombaugh, J., Bittinger, K., Bushman,
F. D., Costello, E. K., … Knight, R. (2010). QIIME allows analysis of
high-throughput community sequencing data. Nature Methods, 7,
335–336. https://doi.org/10.1038/nmeth.f.303
Curd, E. E., Martiny, J. B. H., Li, H., & Smith, T. B . (2018). Bacterial di-
versity is positively correlated with soil heterogeneity. Ecosphere, 9.
https://doi.org/10.10 02/ecs2.2079
De Palma, A., Sanchez-Ortiz, K., Martin, P. A., Chadwick, A., Gilbert, G.,
Bates, A . E., … Purvis, A. (2018). Challenges with inferring how land-
use affects terrestrial biodiversity: Study design, time, space, and
synthesis. Advan ces in Ecological Research, 58, 163–1 9 9.
Delgado-Baquerizo, M., Reich, P. B., Khachane, A. N., C ampbell, C. D.,
Thomas, N., Freitag, T. E., … Singh, B. K. (2017). It is elemental: Soil
nutrient stoichiometry drives bacterial diversity. Environmental
Microbiology, 19, 1176–1188.
Dickens, S. J. M., Allen, E. B., Santiago, L. S., & Crowley, D. (2015).
Extractable nitrogen and microbial community structure respond to
grassland restoration regardless of historical context and soil compo-
sition. AoB Plants, 7. https://doi.org/10.1093/aobpl a/plu085
Dickie, I. A. (2010). Insidious effects of sequencing errors on perceived
diversity in molecular surveys. New Phytologist, 188, 916–918.
https://doi.org/10.1111/j.1469-8137.2010.03473.x
Ding, G.-C., Piceno, Y. M., Heuer, H., Weinert, N., Dohrmann, A. B.,
Carrillo, A., … Smalla, K. (2013). Changes of soil bacterial diversity
as a consequence of land use in a semi-arid ecosystem. PLoS ONE,
8, e59497.
Dong, X. U. E., Huai-Ying, Y. A. O., De-Yong, G. E., & Huang, C.-Y. (2008).
Soil microbial community structure in diverse land use systems: A com-
parative study using Biolog, DGGE, and PLFA analyses. Pedosphere,
18, 653–663. https://doi.org/10.1016/S1002-0160(08)60060-0
Dupouey, J. L., Dambrine, E., Laffite, J. D., & Moares, C. (2002).
Irreversible impact of past land use on forest soils and biodi-
ver si ty. Ecology, 83, 2978–2984. https://doi.org/10.1890/0012-
9658(2002)083[2978:IIOPL U]2.0.CO;2
Edgar, R. C . (2010). Search and clu stering orders of magnitude faster than
BL AST. Bioinformatics, 26 , 2460–2461. https://doi.org/10.1093/bioin
forma tics/btq461
Edgar, R. C. (2013). UPARSE: Highly accurate OTU sequences from mi-
crobial amplicon reads. Nature Methods, 10, 996–998. https://doi.
org/10.1038/nmeth.2604
Edgar, R. C. (2016). UCHIME 2: Improved chimer a prediction for amplico n
sequencing. BioRxiv, 0742 52 .
Edgar, R. C. (2017). Accur acy of microbial community diversity estimated
by closed- and open-reference OTUs. PeerJ, 5, e3889. https://doi.
org /10.7717/peerj.3889
Edgar, R. C ., & Flyvbjerg , H. (2015). Error fi ltering, pair as sembly and erro r
correction for next-generation sequencing reads. Bioinformatics, 31,
3476–3482. https://doi.org/10.1093/bioin forma tics/btv401
Fichtner, A., Von Oheimb, G., Härdtle, W., Wilken, C., & Gutknecht, J.
L. M. (2014). Effects of anthropogenic disturbances on soil micro-
bial communities in oak forests persist for more than 100 years. Soil
Biology and Biochemistry, 70, 79–87. https://doi.org/10.1016/j.soilb io.
2013.12.015
Fierer, N., & Jackson, R. B. (2006). The diversit y and biogeography of
soil bacterial communities. Proceedings of the National Academy of
Sciences of the United States of America, 103, 626–631. https://doi.
org/10.1073/pnas.05075 35103
Flinn, K. M., & Marks, P. L. (2007). Agricultural legacies in forest envi-
ronment s: Tree communities, soil properties, and light availability.
Ecological Applications, 17, 452–463. https://doi.org/10.1890/05-
1963
Flinn, K. M., & Vellend, M. (20 05). Recovery of forest plant communities in
post agricultural landscapes. Frontiers in Ecolog y and the Environme nt,
3, 243–250. https://doi.org/10.1890/1540-9295(2005)003[0243:
ROFPC I]2.0.CO;2
Flinn, K. M., Vellend, M., & Marks, P. L. (2005). Environmental causes
and consequences of forest clearance and agricultural abandonment
in central New York, USA . Journal of Biogeography, 32, 439–452.
https://doi.or g/10.1111/ j.1365-2699.2 004.01198.x
Foley, J. A., DeFries, R ., Asner, G. P., Barford, C., Bonan , G., Carpenter, S.
R., … Snyder, P. K. (2005). Global consequences of land use. Science,
309, 570–574. https://doi.org/10.1126/scien ce.1111772
Foster, D., Swanson, F., Aber, J., Burke, I., Brokaw, N., Tilman, D., &
Knapp, A . (2003). The importance of land-use legacies to ecolog y
and conservation. BioScience, 53, 77–88. https://doi.org/10.1641/
0006-3568(2003)053[0077:TIOLU L]2.0.CO;2
Fox, J., & Weisberg, S. (2018). A n R companion to applied regression.
London, UK: Sage Publications.
Freschet, G. T., Östlund, L., Kichenin, E., & Wardle, D. A. (2014). Above
and belowground legacies of native Sami land use on boreal forest
in northern Sweden 100 years after abandonment. Ecology, 95,
96 3 –9 7 7.
Haegeman, B., Hamelin, J., Moriarty, J., Neal, P., Dushoff, J., & Weitz, J.
S. (2013). Robust estimation of microbial diversity in theory and in
practice. The ISME Journal, 7, 1092–1101. https://doi.org/10.1038/
ismej.2013.10
Hahn, P. G., & Orrock, J. L. (2015). Spatial arrangement of canopy struc-
ture and land-use history alter the effect that herbivores have
on plant growth. Ecosphere, 6, 193. https://doi.org/10.1890/ES15-
00036.1
Harris, J. (2009). Soil microbial communities and restoration ecol-
ogy: Facilitators or followers? Science, 325, 573–574. https://doi.
org/10.1126/scien ce.1172975
Hartman, W. H., Richardson, C . J., Vilgalys, R., & Bruland, G. L. (2008).
Environmental and anthropogenic controls over bacterial communi-
ties in wetland soils. Proceedings of the National Academy of Science s
of the United States of America, 105, 17842–17847. https://doi.
org/10.1073/pnas.08082 54105
Hermy, M., & Verheyen, K. ( 2007). Legacie s of the past in the present-day
forest biodiversity: A review of past land-use effects on forest plant
species composition and diversity. Ecological Research, 22, 361–371.
https://doi.org/10.1007/s11284-007-0354-3
Hui, N., Liu, X., Jumpp onen, A., Set älä, H., Kotze, D. J., Biktasheva, L.,
& Romantschuk, M. (2018). Over twenty years farmland reforesta-
tion decreases fungal diversity of soils, but stimulates the return of

862
|
Journal of Applied Ecology
TURLE Y ET aL.
ectomycorrhizal fungal communities. Plant and Soil, 427, 231–244.
https://doi.org/10.1007/s11104-018-3647-0
Jaeger, B. (2017). r2glmm: Computes R squared for mixed (multilevel)
models. R package version 0.1.2. Retrieved from https://CRAN.R-
proje ct.org/packa ge=r2glmm
Jangind, K., W illiams, M. A ., Franzluebbers, A. J., Schmidt, T. M.,
Coleman, D. C., & Whitman, W. B. (2011). Land-use history has a
stronger impact on soil microbial community composition than abo-
veground vegetation and soil properties. Soil Biology & Biochemistry,
43, 2184–2193. https://doi.org/10.1016/j.soilb io.2011.06.022
Jesus, E. D. C., Marsh, T. L., T iedje, J. M., & Moreira, F. M. D. S. (20 09).
Changes in land use alter the structure of bacterial communities in
Western Amazon soils. The ISME Journal, 3, 1004–1011. https://doi.
org /10.1038/ismej.2009.47
Jones, H. P., Jones, P. C., Barbier, E. B., Blackburn , R. C., Rey Benayas, J.
M., Holl, K. D., … Moreno Mateos, D. (2018). Restoration and repair
of Earth's damaged ecosystems. Pro ceedings of the Royal Society B:
Biological Sciences, 285(1873), 2017–2577. https://doi.or g/10 .1098/
rspb.2017.2577
Kardol , P., Martijn Bezemer, T., & Van Der Putten, W. H. (20 06). Temporal
variation in plant–soil feedback controls succession. Ecology Letters,
9, 1080–1088 . https://doi.org/10.1111/j.1461-024 8. 2006.0 0953. x
Kardol, P., & Wardle, D. A. (2010). How understanding aboveground-
belowground linkages can assist restoration ecology. Trends in Ecolog y
& Evolution, 25, 670–679. https://doi.org/10.1016/j.tree.2010.09.001
Keiser, A. D., Strickland, M. S., Fierer, N., & Bradford, M. A. (2011). The
effect of resource history on the functioning of soil microbial com-
munities is maintained across time. Biogeosciences, 8(6), 1477–1486 .
https://doi.org/10.5194/bg-8-1477-2011
Kilgo, J. C ., & Blake, J. I. (20 05). Ecology and management of a forested
landscape: Fifty years on the Savannah River Site. Washington, DC:
Island Press.
Kõljalg, U., Nilsson, R. H., Abarenkov, K., Tedersoo, L., Taylor, A. F. S.,
Bahram, M., … Larsson, K.-H. (2013). Towards a unified paradigm for
sequence-based identification of fungi. Molecular Ecology, 22, 5271–
5277. ht tps://doi.org /10.1111/me c.12481
Koziol, L., Schultz, P. A., House, G. L., Bauer, J. T., Middleton, E. L ., &
Bever, J. D. (2018). The plant microbiome and native plant resto-
ration: The example of native mycorrhizal fungi. BioScience, 68, 996–
1006. https://doi.org/10.1093/biosc i/biy125
Lauber, C. L., Strickland, M. S., Bradford, M. A ., & Fierer, N. (2008). The
influence of soil properties on the structure of bacterial and fungal
communities across land-use types. Soil Biology & Biochemistry, 40,
2407–2415. https://doi.org/10.1016/j.soilb io.2008.05.021
Love, M. I., Huber, W., & Anders, S. (2014). Moderated estimation of
fold change and dispersion for RNA-seq data with DESeq2. Genome
Biology, 15, 550. https://doi.org/10.1186/s13059-014-0550-8
Ma, S., De Frenne, P., Boon, N., Brunet, J., Cousins, S. A. O., Decocq, G.,
… Verheyen, K. (2019). Plant species identity and soil characteris-
tics determine rhizosphere soil bacteria community composition in
European temperate forests. FEMS Microbiology Ecology, 95, fiz063.
https://doi.org/10.1093/femse c/fiz063
Ma, S., De Frenne, P., Vanhellemont, M., Wasof, S., Boeckx, P., Brunet,
J., … Verheyen, K. (2019). Local soil characteristics determine the
microbial communities under forest understorey plants along a lat-
itudinal gradient. Basic and Applied Ecology, 36, 34–44. https://doi.
org/10.1016/j.baae.2019.03.001
Ma, S., Verheyen, K., Props, R., Wasof, S., Vanhellemont, M., Boeckx, P.,
… De Frenne, P. (2018). Plant and soil microbe responses to light,
warming and nitrogen addition in a temperate forest. Functional
Ecology, 32, 1293–1303. ht tps://doi.or g/10.1111/1365-2435.13061
Masella, A . P., Bartram, A . K., Truszkowski, J. M., Brown, D. G ., & Neufeld,
J. D. (2012). PANDAseq: Paired-end assembler for illumina sequences.
BMC Bioinformatics, 13, 31. https://doi.org /10.1186/1471-2105-
13-31
Meli, P., Holl, K. D., Rey Benayas, J. M., Jones, H. P., Jones, P. C ., Montoya,
D., & Mateos, D. M. (2017). A global review of past land use, climate,
and active vs. passive restoration effects on forest recovery. PLoS
ONE, 12, e0171368.
Newbold, T., Hudson, L. N., Hill, S. L. L., Contu, S., Lysenko, I., Senior,
R. A., … Purvis, A. (2015). Global effects of land use on local terres-
trial biodiversity. Nature, 520, 45–50. https://doi.org/10.1038/natur
e14324
Noss, R. F., Platt, W. J., Sorrie, B. A., Weakley, A. S., Means, D. B., Costanza,
J., & Peet, R. K. (2015). How global biodiver sity hotspot s may go un-
recognized: Lessons from the North American coastal plain. Diversity
and Distributions, 21, 236–244. https://doi.org/10.1111/ddi.12278
Oehl, F., Sieverding, E., Ineichen, K., Mäder, P., Boller, T., & Wiemken, A.
(2003). Impact of land use intensit y on the species diversity of arbus-
cular mycorrhizal fungi in agroecosystems of Central Europe. Applied
Environmental Microbiology, 69, 2816–2824. https://doi.org/10.1128/
AEM.69.5.2816-2824.2003
Oksanen, J., Blanchet, F. G., Kindt , R., Legendre, P., O’hara, R. B., Simpson,
G. L., … Wagner, H. (2010). Vegan: community ecology package.
R package version 1.17-4. Retrieved from http://cran.r-proje ct.org>
Peay, K. G., Baraloto, C., & Fine, P. V. A. (2013). Strong coupling of plant
and fungal community structure across western Amazonian rainforests.
The ISME Journal, 7, 1852–1861. https://doi.org/10.1038/ismej.2013.66
Perring, M. P., De Frenne, P., Baete n, L., Maes, S. L ., Depauw, L., Blondeel,
H., … Verheyen, K. (2016). Global environmental change effects on
ecosystems: The importance of land-use legacies. Global Change
Biology, 22, 1361–1371. h tt ps://doi.or g/10.1111/gcb.13146
Perring, M. P., Standish, R. J., Price, J. N., Craig, M. D., Erickson, T. E.,
Ruthrof, K. X., … Hobbs, R . J. (2015). Advances in restoration ecol-
ogy: Rising to the challenges of the coming decades. Ecosphere, 6,
131. https://doi.org/10.1890/ES15-00121.1
Potthof f, M., Steenwerth, K. L ., Jackson, L . E., Drenovsk y, R. E., Scow, K.
M., & Joergensen, R. G. (2006). Soil microbial community composi-
tion as affected by restoration practices in California grassland. Soil
Biology and Biochemistry, 38, 1851–1860. https://doi.org/10.1016/
j.soilb io.2005.12.009
Prober, S. M., Leff, J. W., Bates, S. T., Borer, E. T., Firn, J., Harpole, W. S.,
… Fierer, N. (2015). Plant diversity predicts beta but not alpha diver-
sity of soil microbes across grasslands worldwide. Ecology Letters, 18,
85–95. ht tps://doi.org/10.1111/ele.12381
Rodrigues, J. L. M., Pellizari, V. H., Mueller, R., Baek, K ., Jesus, E. D. C.,
Paula, F. S., … Nusslein, K. (2013). Conversion of the Amazon rain-
forest to agriculture results in biotic homogenization of soil bacte-
rial communities. Proceedings of the National Academy of S ciences of
the United States of America, 110, 988–993. https://doi.org/10.1073/
pnas.12206 08110
Rosseel, Y. (2012). lavaan: An R package for structural equation model-
ing. Journal of Statistical Sof tware, 48(2), 1–36. Retrieved from http://
www.jstat soft.org/v48/i02/
Semenova-Nelsen, T. A., Platt, W. J., Patterson, T. R., Huffman, J., & Sikes,
B. A. (2019). Frequent fire reorganizes fungal communities and slows
decomposition across a heterogeneous pine savanna landscape. New
Phytologist, 224, 916–927. htt ps://doi.or g/10.1111/n ph .16096
Strickland, M. S., Callaham Jr., M. A ., Gardiner, E. S., Stanturf, J. A., Leff,
J. W., Fierer, N., & Bradford, M. A. (2017). Response of soil microbial
community composition and function to a bottomland forest resto-
ration intensity gradient. Applied Soil Ecology, 119, 317–326. https://
doi.org/10.1016/j.apsoil.2017.07.008
Stuhler, J. D., & Orrock, J. L. (2016). Historic al land use and present-day
canopy thinning differentially affect the distribution and abun-
dance of invasive and native ant species. Biological Invasions, 18,
1813–1825. https://doi.org/10.1007/s10530-016-1122-5
Suding, K. N. (2011). Toward and era of restoration ecology: Successes,
failures, and opportunities ahead. Annual Revi ew of Ecology, Evolution,
and Systematics, 42, 465–487.

|
863
Journal of Applied Ecolog
y
TURLE Y ET aL.
Thiéry, O., Moora, M., Vasar, M., Zobel, M., & Öpik, M. (2012). Inter- and
intrasporal nuclear ribosomal gene sequence variation within one
isolate of arbuscular mycorrhizal fungus, Diversispora sp. Sy mbiosis,
58, 135–147. https://doi.org/10.1007/s13199-012-0212-0
Turley, N. E., & Brudvig, L. A. (2016). Agricultural land-use history causes
persistent loss of plant phylogenetic diversity. Ecology, 97, 2 240–
2247. https://doi.org/10.1002/ecy.1443
Turley, N., Brudvig, L., Bell-Dereske, L ., & Evans, S. (2020). Data from:
Agricultural land-use history and restoration impact soil microbial
biodiversity. Dryad Digital Repositor y, ht tps://do i.org /10.50 61/drya d.
x3ffb g7fd
Turley, N. E., Orrock, J. L., Ledvina, J. A., & Brudvig , L. A. (2017). Dispersal
and establishment limitation slows plant community recovery in
post-agricultural longleaf pine savannas. Journal of Applied Ecology,
54, 1100–1109.
Upchurch, R., Chiu, C. Y., Everett, K., Dyszynski, G., Coleman, D. C., &
Whitman, W. B. (2008). Differences in the composition and diver-
sity of bacterial communities from agricultural and forest soils. Soil
Biology and Biochemistry, 40, 1294–1305.
van der Bij, A. U., Weijters , M. J., Bobbink, R., Harris, J. A., Pawlett,
M., Ritz, K., … van Diggelen, R. (2018). Facilitating ecosystem
assembly: Plant-soil interactions as a restoration tool. Biological
Conservation, 220, 272–279. https://doi.org/10.1016/j.biocon.2018.
02.010
van der Heijden, M. G. A., Bardgett, R. D., & van Straalen, N. M. (20 08).
The unseen majority: Soil microbes as drivers of plant diversity and
productivity in terrestrial ecosystems. Ecology Letters, 11, 296–310.
https://doi.or g/10.1111/ j.14 61-0248.2007.01139.x
Wagg, C., Dudenhöffer, J. H., Widmer, F., & Van Der Heijden, M. G.
(2018). Linking diversity, synchrony and stability in soil micro-
bial communities. Functional Ecology, 32, 1280–1292. https://doi.
org /10.1111/1365-2435.13056
Weiss, S., Xu, Z. Z., Peddada, S., Amir, A., Bittinger, K., Gonzalez, A., …
Knight, R. (2017). Normalization and microbial differential abun-
dance strategies depend upon data characteristics. Microbiome, 5,
27. https://doi.org/10.1186/s40168-017-0237-y
Wubs, E. R. J., van der Putten, W. H., Bosch, M., & Bezemer, T. M. (2016).
Soil inoculation steers restoration of terrestrial ecosystems. Nature
Plants, 2, 16107. https://doi.org/10.1038/nplan ts. 2016 .107
Xue, P. P., Carrillo, Y., Pino, V., Minasny, B., & McBratney, A. B. (2018).
Soil properties drive microbial community structure in a large scale
transect in south eastern Australia. Scientific Reports, 8, 11725.
SUPPORTING INFORMATION
Additional supporting information may be found online in the
Supporting Information section.
How to cite this article: Turley NE, Bell-Dereske L, Evans SE,
Brudvig LA. Agricultural land-use history and restoration
impact soil microbial biodiversity. J Appl Ecol. 2020;57:852–
863. ht tps://doi.org/10.1111/1365-266 4.13591