ArticlePDF Available

Garden soil bacteria transiently colonize gardeners' skin after direct soil contact

Wiley
Urban Agriculture & Regional Food Systems
Authors:

Abstract and Figures

Urban soils provide a number of ecosystem services and health benefits, yet they are understudied compared with agricultural and wildland soils. Healthy soils host diverse microbiota, exposure to which may be critical for immune development and protection against chronic disorders, such as allergies and asthma. Gardening represents a key pathway for microbiota exposure, yet little is known about microbial community structure of urban garden soils, degree of soil‐to‐skin transfer during gardening, nor ability of soil microbes to persist on human skin. To explore these questions, we recruited 40 volunteers to collect soil samples from their gardens and a series of skin swab samples before and after gardening. Soil and skin bacterial communities were characterized using amplicon (16S) sequencing. Soil samples were also analyzed for chemical/physical characteristics. Soil bacterial communities had more alpha diversity and less beta diversity than skin communities, which varied greatly across individuals and within the same individual across time. The number of bacterial taxa shared between skin and garden soil increased immediately after gardening for most study participants. However, the imprint of garden soil largely disappeared within 12 hours. Despite this lack of persistence, a daily gardening routine with repeated and extended contact with soil likely reinoculates the skin such that soil microbes are often present, holding potential to impact health.
This content is subject to copyright. Terms and conditions apply.
Received: 10 August 2022 Accepted: 2 December 2023
DOI: 10.1002/uar2.20035
ORIGINAL ARTICLE
Special Section: Improving Livability in Urban Areas: Examining Urban and Peri-Urban Soil and Plant
Management
Garden soil bacteria transiently colonize gardeners’ skin after
direct soil contact
Gwynne Á. Mhuireach1Kevin G. Van Den Wymelenberg1Gail A. Langellotto2
1Biology and the Built Environment Center,
University of Oregon, Eugene, OR, USA
2Department of Horticulture, Oregon State
University, Corvallis, OR, USA
Correspondence
Gwynne Á. Mhuireach, Biology and the
Built Environment Center, University of
Oregon, Eugene, OR, USA.
Email: gwynhwyf@uoregon.edu
Abstract
Urban soils provide a number of ecosystem services and health benefits, yet they
are understudied compared with agricultural and wildland soils. Healthy soils host
diverse microbiota, exposure to which may be critical for immune development and
protection against chronic disorders, such as allergies and asthma. Gardening rep-
resents a key pathway for microbiota exposure, yet little is known about microbial
community structure of urban garden soils, degree of soil-to-skin transfer during
gardening, nor ability of soil microbes to persist on human skin. To explore these
questions, we recruited 40 volunteers to collect soil samples from their gardens and a
series of skin swab samples before and after gardening. Soil and skin bacterial com-
munities were characterized using amplicon (16S) sequencing. Soil samples were
also analyzed for chemical/physical characteristics. Soil bacterial communities had
more alpha diversity and less beta diversity than skin communities, which varied
greatly across individuals and within the same individual across time. The number
of bacterial taxa shared between skin and garden soil increased immediately after
gardening for most study participants. However, the imprint of garden soil largely
disappeared within 12 hours. Despite this lack of persistence, a daily gardening rou-
tine with repeated and extended contact with soil likely reinoculates the skin such
that soil microbes are often present, holding potential to impact health.
1INTRODUCTION
Soil health has been defined relative to a soil’s ability to
support plant and animal health and productivity, as well
as water and air quality, within the bounds and context set
by ecosystem and climate (Doran & Safley, 1997; Doran,
2002). Urban soils, as a subset of global soil types, are under-
studied and their role in providing ecosystem services, such
as climate regulation (e.g., carbon sequestration), biocon-
trol, nutrient cycling, and soil formation (Kibblewhite et al.,
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and distribution in any medium, provided
the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
©2023 The Authors. Urban Agriculture & Regional Food Systems published by Wiley Periodicals LLC on behalf of American Society of Agronomy and Crop Science Society of
America.
2008; Lehmann et al., 2020; Norris et al., 2020), has been
long underappreciated (O’Riordan et al., 2021). Furthermore,
urban soil health has been identified as a key considera-
tion of city planning and soil science (Guilland et al., 2018;
Adewopo et al., 2014), in large part because of the grow-
ing importance of gardens and other types of greenspace to
urban sustainability and resilience (McPhearson et al., 2014).
In addition to the aforementioned regulating and supporting
services, urban soils found in community and private gar-
dens also provide provisioning (e.g., food, fiber, and medicine
Urban Agric Region Food Syst. 2023;8:e20035. wileyonlinelibrary.com/journal/uar2 1of22
https://doi.org/10.1002/uar2.20035
2of22 MHUIREACH ET AL.
production) and social (e.g., mental/physical health benefits
and opportunity for social interaction) services (O’Riordan
et al., 2021; Teuber, 2021; Schram-Bijkerk et al., 2018;
Rook, 2013; Taylor & Lovell, 2015,2014). Importantly, many
ecosystem services provided by urban gardens are inter-
connected and synergistic, leading to simultaneous benefits
for environmental and human health (Schram-Bijkerk et al.,
2018). However, research questions related to the relationship
between garden soil health and human health are complicated
by the heterogeneous nature of urban soils, which is some-
times referred to as the “urban soil mosaic” (Pouyat et al.,
2010). Decisions gardeners make when building and tending
garden soils have a strong influence on soil health (Tresch
et al., 2018) and often result in garden soils that are sub-
stantially dissimilar to parent native soils. Therefore, different
considerations and research questions drive our understand-
ing of urban garden soils, compared to rural agricultural or
wildland soils (Nelson et al., 2022).
Microbes play an essential role in realization of ecosys-
tem services, via biological activities associated with soil
aggregate stabilization, organic matter decomposition, con-
trol of soil pathogens, crop production, and even human
immune system development (Kibblewhite et al., 2008).
Given their established importance to soil structure and func-
tion, microbes are often used as bioindicators of soil health
(Fierer et al., 2021). Agricultural intensification is associ-
ated with declines in soil health, generally (Tilman et al.,
2002), and declines in microbial communities, specifically
(Banerjee et al., 2019). Management practices, including
differences in organic versus conventional agricultural sys-
tems, are key drivers of soil microbiota (Araújo et al., 2009;
Hartmann et al., 2015). For instance, mechanical cultivation
and continuous crop production in conventional farming sys-
tems result in physical soil loss through erosion and decreases
in soil organic matter (Naylor, 1996), which also disrupt and
damage microbial communities (Chen et al., 2020). Organic
systems, by comparison, tend to have higher soil organic car-
bon and greater microbial activity and biomass (Henneron
et al., 2015; Lori et al., 2017; Martìnez-Garcìa et al., 2018;
Verbruggen et al., 2010), although these qualities may vary
within organic systems depending on specific farm character-
istics and practices (Lupatini et al., 2017; Verbruggen et al.,
2010; Dumontet et al., 2017). Manure application, in par-
ticular, tends to increase soil microbial biomass and activity
(Esperschütz et al., 2007; van der Bom et al., 2018;Widmer
et al., 2006). Variation in plant species composition may also
affect soil microbial community diversity and composition
near the root zone, since many plants produce root exudates
that can encourage beneficial microorganisms or discourage
pathogens (Lei et al., 2019; Schmid et al., 2019).
A caveat to these findings is that the majority of studies
have been conducted in nonurban agricultural fields, although
one recent study of urban soils across 85 gardens in Zurich,
Core Ideas
Garden soil microbiomes were diverse but rela-
tively homogeneous across geographic locations
and climate zones.
Skin microbiomes exhibited low diversity and high
inter- and intrapersonal variation.
After gardening, the number of bacterial taxa
shared between gardeners’ skin and their garden
soil increased.
Most soil-associated microorganisms were tran-
sient on the skin and did not persist beyond 12
hours.
Switzerland, found that organic versus conventional man-
agement practices are important factors driving soil health
and function (Tresch et al., 2018). Gardens are perhaps the
most ubiquitous form of urban agriculture (Taylor & Lovell,
2014; McClintock et al., 2013), thus it is important that we
understand how gardening activities and management prac-
tices influence soil characteristics and associated ecosystem
services. Reducing tillage, applying compost, and increas-
ing plant diversity have been suggested as ways to improve
urban garden soil microbial community structure and function
(Salomon & Cavagnaro, 2022).
Soil health and biodiversity are closely linked with human
health. Many health-relevant roles played by soil have been
extensively investigated, such as the ability to produce
nutritious food, sequester carbon, remediate environmental
pollutants, and purify water. In recent years, increased atten-
tion has been given to the potential role that soil may have on
human health via impacts on the human microbiome. Some
of these studies focus on the indirect role that soil can play on
human gut microbiota, via agricultural management practices
that ultimately impact soil microbes on the fruits and vegeta-
bles that we consume (Blum et al., 2019; Hirt, 2020). Until
recently, research into direct links between urban soil and
human health focused on detrimental effects following from
ingestion of heavy metals, other toxins, pathogens, or para-
sites in soil (Oliver, 1997; Abrahams, 2002). However, new
evidence suggests that exposure to environmental microbiota,
including those associated with soil, may be important for
establishment of normal gut flora and healthy immune system
development (von Hertzen & Haahtela, 2006; Schmidt et al.,
2011; Mulder et al., 2009), particularly if soil biodiversity is
robust (Wall et al., 2015).
Because humans evolved for millennia in the presence
of these environmental microbes associated with vegetation,
soil, water, and wildlife, our immune systems are not only
adapted to coexist with the majority of these microbes, but
MHUIREACH ET AL.3of22
may even require that interaction to function properly (Rook,
2021). Due to technological advances in human microbiome
studies using next-generation DNA sequencing, we now
know that microbes are critical for an array of mammalian
biological processes, including metabolism, brain develop-
ment, and immune system responses (Rook et al., 2017;
McFall-Ngai et al., 2013). Furthermore, a number of inflam-
matory diseases are tied to microbial imbalances, or dysbiosis,
possibly related to increasingly urban lifestyles (Blaser &
Falkow, 2009;Blaser,2014). Several related hypotheses—
the Old Friends Hypothesis (Rook et al., 2003; Rook, 2021),
Biodiversity Hypothesis (Haahtela, 2019), and Microbiome
Rewilding Hypothesis (Mills et al., 2017) —have suggested
that exposure to the natural world and, especially, biodiverse
microbiota in natural spaces, can reduce disease. Several cor-
relative examples are cited in support of these hypotheses,
including the higher incidence of chronic inflammatory dis-
orders (e.g., asthma and allergies) in urban populations with
limited access to green spaces (Bach, 2002; Rook et al., 2017;
von Hertzen et al., 2011; Hanski et al., 2012; McDade, 2012)
and the protective effect of traditional farming environments
with a high degree of diverse microbial exposures (Stein et al.,
2016). Supporting this idea, children living in more biodiverse
areas, such as farms or forests, tend to have greater diversity
of Gammaproteobacteria on their skin and lower prevalence
of allergies and asthma than their urban counterparts (Hanski
et al., 2012; Ruokolainen et al., 2015). Additionally, a recent
intervention experiment demonstrated that the biodiversity
of children’s play-yards is positively associated with skin
Gammaproteobacteria diversity and immunoregulatory func-
tion (Roslund et al., 2020). It is currently unknown whether
these health effects are related to specific microorganisms,
microbial consortia, microbial biodiversity in a broader sense,
or a combination, as current evidence supports all three mech-
anisms (Matthews & Jenks, 2013; Ege et al., 2011; Hanski
et al., 2012).
Noted associations between urban living and inflamma-
tory disease are particularly troubling, in the context of an
increasingly urbanized world. On a global basis, 55% of
the world’s population is urban, with 82% of North Amer-
icans living in urban versus rural areas (United Nations,
2019). Within urban areas, access to greenspaces is often dis-
proportionately restricted according to socio-economic level
(Kuras et al., 2020; Leong et al., 2018). In addition to less
access to greenspace, lower income households also fre-
quently experience higher incidence of inflammatory disease
(Rook et al., 2014), which some have argued reframes envi-
ronmental microbial exposure as a social equity issue (Ishaq
et al., 2019).
Gardening provides an accessible means for urban residents
to interact with soils, plants, and environmental microbiota.
Because most cultivated vegetables are grown as annual
plants, vegetable gardening provides repeated opportunities
to interact with the soil, compared to cultivation of perennial,
ornamental plants. A single gram of garden soil may con-
tain billions of individual microorganisms, including bacteria,
fungi, viruses, and archaea, representing thousands of differ-
ent microbial species (Torsvik & Øvreås, 2002). Subungual
regions (under the fingernails) have previously been shown
to act as a reservoir of large numbers of microorganisms,
in comparison to other parts of the hand (McGinley et al.,
1988). We speculate that this reservoir may be particularly
sizable when soil particles are trapped under the fingernail,
as may frequently occur in the context of gardening. Despite
the likelihood of substantial exposure to soil microbes during
gardening activities, little is known about how much microbial
transfer from soil to skin occurs, what types of microorgan-
isms are transferred, how long they can persist, or how they
might affect human health. Recent studies have shown that
direct contact with soil and its associated microbiota can leave
an imprint on the skin microbiome for at least 24 h, even after
washing and bathing (Grönroos et al., 2018). In addition, soil-
associated taxa that are acquired during gardening can later be
transferred to the human gut (Nurminen et al., 2018;Brown
et al., 2022). This process may have evolutionary prece-
dent, as anthropological research has unearthed evidence that
human population groups with more intimate relationships
with nature (e.g., rural farmers and hunter-gatherers) tend to
accumulate more soil-associated microbiota in their guts than
urban dwellers (Schnorr, 2020). At the same time, the poten-
tial for food contamination by pathogenic microbes in urban
soils has been highlighted as a critical issue requiring study
(Hallett et al., 2016; Salomon & Cavagnaro, 2022).
The primary goal of this study, therefore, was to advance
scientific understanding of soil-to-skin microbial transfer
dynamics as a result of direct soil contact associated with
gardening. Our research objectives were to: (1) determine
how many (relative abundance) soil microbes are trans-
ferred to skin during gardening, which taxa are transferred,
and how long they persist during normal hygiene routines;
(2) investigate whether management practices, crop type,
and/or location influence soil microbial community diversity
and composition; and (3) test whether effects on the skin
microbiome vary by management practices or location.
2METHODS
2.1 Gardener recruitment
We recruited study volunteers by posting on the Oregon
State University Garden Ecology Lab blog on June 4,
2020, and through an informational webinar on June 5,
2020 (https://blogs.oregonstate.edu/gardenecologylab/2020/
06/05/link-to-gardener-microbiome-webinar-recording/).
Respondents were screened using a standard screening
4of22 MHUIREACH ET AL.
FIGURE 1 Map of general study site locations within Oregon (USA), colored by management practices. A geographic masking technique has
been applied to protect the privacy of study participants.
questionnaire (Supplemental File 1). To be eligible for the
study, volunteers were required to be over the age of 18,
fluent in English, involved in regular gardening activities,
and not have any health conditions that would limit direct
contact with soil. There were 66 total respondents to our
recruitment program, from which we selected 40 participants
representing equal numbers from the Willamette Valley and
High Desert regions (Figure 1), and equal numbers reporting
organic and nonorganic management practices. Although
we did not include urban location as a participation criteria
in the initial screening survey, over 75% of the gardens
from which data were collected were located in urban areas,
according to the 2010 US Census Bureau definitions. Two
participants withdrew prior to collecting samples and two
additional gardeners representing the appropriate locations
and management practices were selected from the recruitment
list as replacements. Note that sampling kit numbers were not
altered to account for dropouts and replacements, such that
kit numbers reported in the results section (e.g., Figures 5
and 9) are not entirely consecutive.
This study was approved by the University of Oregon
Institutional Review Board on May 5, 2020, (Approval No.
03112020.013), and all participants provided informed con-
sent. Geographic masking techniques (Zandbergen, 2014)
have been applied on all published and/or publicly accessible
maps to protect the privacy of study participants.
2.2 Sample collection
Soil and skin microbiome samples were self-collected by
study participants during July–September, 2020. Prior to
collecting samples, participants received training in collec-
tion protocols via an online video and written instruction
packet (Supplemental File 2). Individual sampling kits were
provided to participants, each containing three resealable
plastic bags for soil samples, four sterile cotton-tipped swabs
(Puritan #25-806) for skin surface sample collection, and
four microcentrifuge tubes containing phosphate-buffered
saline solution. Each participant collected soil samples from
MHUIREACH ET AL.5of22
three different garden beds, noting crop type(s) planted
within each bed. Each soil sample comprised three smaller
subsamples that were combined in the same resealable plastic
bag. Participants were asked to scrape aside any mulch or
top-dressing layer and use a trowel or small hand shovel to
collect roughly 1 cup (approximately 250–300 g) for each
subsample, digging to a depth of 10 cm. Participants were
asked to collect skin microbiome swabs from their hand (dor-
sal surface, interdigital spaces, fingernails, and subungual
regions) at four timepoints: before, immediately after, 12 h
after, and 24 h after gardening. Swabs were moistened in
the phosphate-buffered saline buffer solution prior to sample
collection, and placed in the remaining buffer in the micro-
centrifuge tube after collection. Participants were asked to
avoid contact with soil for a 24-h period prior to the first swab
collection, to ensure soil contact through normal gardening
activities for at least 2 h on the first day of sampling, and to
avoid additional soil contact for the following 24-h period.
Soil and skin swab samples were stored in participants’
home freezers until being mailed to the Biology and Built
Environment (BioBE) Center at University of Oregon (UO).
Study participants were also asked to complete two ques-
tionnaires (Supplemental Files 3and 4): one with questions
about garden management practices, results of which are
compiled in a crosstab (Table S1), and the other containing
questions about skin care and lifestyle characteristics (e.g.,
topical antibiotic use, presence of indoor pets, hand-washing
frequency, emollient application) that could influence base-
line skin microbiome composition (Kong et al., 2017). We did
not expect skin care and/or lifestyle characteristics to affect
potential transfer of soil microorganisms, although they could
conceivably impact persistence. Participants mailed their soil
and skin microbial samples and questionnaires to the BioBE
Center where microbial samples were immediately stored at
-80C until further processing.
2.3 Bacterial processing and analysis
Soil samples were homogenized and sieved with a 6.35 mm
mesh sieve. Each sample was aliquoted into two 0.25 g sub-
samples(technicalreplicates)anda1gsubsampletobestored
as a backup. Two technical replicates from each soil core
were prepared for sequencing, as single samples typically
capture only a small fraction of the microbial taxa actually
present (Castle et al., 2019). The remaining bulk soil was
sent to Oregon State University (OSU) Soil Health Lab for
physical and chemical assays. We used Earth Microbiome
Project protocols (Thompson et al., 2017) for all laboratory
processing steps. Soil-specific protocols and primers were
chosen, despite their limited ability to discriminate certain
skin-associated taxa, because our primary research questions
revolved around the transfer and fate of soil microorganisms
on skin, rather than the skin microbiome per se. In brief, we
extracted genomic DNA from soil subsamples and skin swabs
using the Qiagen DNeasy PowerSoil Kit, following manu-
facturer’s instructions. After extraction, bacterial DNA was
amplified in triplicate using universal primer pair 515F–806R
(Caporaso et al., 2012) to target the V4 region of the bac-
terial 16S rRNA gene, then pooled, cleaned, and quantified
prior to sequencing. Samples were sent to the UO Genomics
Core & Cell Characterization Facility for sequencing on an
Illumina MiSeq 2 ×250 PE platform. To control for poten-
tial laboratory contaminants and batch effects, we included
negative controls at all steps and randomized samples during
extraction, PCR, and across the two sequencing runs.
Bioinformatic processing of raw 16S reads was performed
in R version 4.1.0 (R Development Core Team, 2010)using
the DADA2 package for filtering, merging, cleaning, and
assigning taxonomy to paired-end reads (Callahan et al.,
2016). Based on inspection of forward and reverse read qual-
ity plots, we did not truncate reads and we set the maximum
expected error scores to 2 and 3, respectively. Amplicon
sequence variants (ASVs) resulting from the DADA2 pipeline
were assigned taxonomy down to genus level using the Ribo-
somal Database Project (RDP) Bayesian Classifier (Wang
et al., 2007).
We used the R package decontam to identify likely con-
taminants (Davis et al., 2018), which we removed, resulting in
elimination of 2,871,542 reads. A total of five samples failed
to meet our minimum read threshold of 1000 reads and were
removed. Finally, control samples and ASVs representing
non-target organisms (e.g., mitochondria and chloroplasts)
were also removed and technical replicates for each soil
sample were combined prior to downstream analysis.
2.4 Analysis of soil physical/chemical
attributes
Remaining bulk soil from each sample was submitted to the
OSU Soil Health Lab, for Western (Willamette Valley sam-
ples) or Eastern (High Desert samples) Oregon basic analysis,
which included assays for pH, ammonium nitrogen (NH4-
N), nitrate nitrogen (NO3-N), phosphate (PO4-P), potassium
(K), magnesium (Mg), calcium (Ca), and carbon (C). Percent
organic matter (OM) was determined by doubling soil carbon
content. NO3-N, P, K, Mg, Ca are reported as parts per mil-
lion (ppm); NH4-N values were not used in statistical testing,
since ammonium readily converts to nitrate and samples were
collected midseason, when fertilizers were not expected to be
applied. The laboratory used standard analytical procedures to
assess soil parameters, including Moebius-Clune et al. (2016)
for soil pH, Mehlich-III (M3) for P, K, Ca, and Mg, mod-
ifications to Tiessen et al. (1981) for total organic matter,
and KCl extraction for NO3-N.For phosphate, the Bray 01
6of22 MHUIREACH ET AL.
extraction method was used for Willamette Valley samples
and the Olsen sodium bicarbonate extraction method was used
for High Desert samples.
2.5 Statistical analysis
All statistical analyses and data visualizations were also per-
formed in R using packages phyloseq (McMurdie & Holmes,
2013) and ggplot2 (Wickham, 2016) for data wrangling and
visualization. To adjust for unequal bacterial library sizes,
we used the variance-stabilizing transformation in DESeq2,
which models read counts based on a gamma-Poisson, or neg-
ative binomial, distribution (Love et al., 2014). The resulting
transformed counts were used for all microbiome statistical
analyses, except alpha diversity. We set an a priori alpha
threshold of 0.05 and controlled for false discovery rate
inflation using the Benjamini–Hochberg method.
2.5.1 Soil physical/chemical attributes
With the exception of pH, for which residuals were nor-
mally distributed, all soil physical and chemical attributes
were transformed prior to statistical testing. Box–Cox trans-
formation was used for OM, NO3-N, PO4-P, and Ca, while
log transformation was used for K and Mg. Welch’s t-test
was used to evaluate differences in group means according
to management practices and geographic location.
2.5.2 Bacterial alpha diversity
Alpha diversity, or the number of different species observed
within a sample, was quantified as observed richness and
effective number of species, which equates to the num-
ber of equally abundant species needed to achieve a given
alpha diversity metric value, based on Shannon entropy and
Gini–Simpson index using iNEXT (Hsieh et al., 2016; Chao
et al., 2014). These three alpha diversity metrics represent
Hill numbers 0, 1, and 2, respectively, which give varying
degrees of weight to evenness of species distribution, with
species richness giving the least weight to evenness, Gini–
Simpson giving the most, and Shannon entropy falling in
between. For statistical testing, only effective numbers of
species based on Shannon entropy were used, as this met-
ric outperforms others when community weights are unequal
(Jost, 2007), which is commonly the case in microbial com-
munities. Rarefaction/extrapolation curves were examined to
assess sample completeness (Deng et al., 2015; Gotelli & Col-
well, 2001). Prior to performing formal statistical testing, we
explored the data using boxplots to evaluate potential asso-
ciations between fine-grained management practices (e.g.,
tillage, use of organic/synthetic pesticides, and addition of
compost or manure), as reported in the participant garden
management questionnaires, and soil microbial community
structure. Ultimately, we judged that including the fine-
grained management practices could introduce problems with
multicollinearity and would not provide substantial additional
benefit for the analyses.
We used the Mann– Whitney U test to assess differences in
alpha diversity between soil and skin samples. To avoid pseu-
doreplication, we used the mean effective number of species
for soil samples from the same garden and for skin samples
from the same individual.
To examine drivers of garden soil alpha diversity variation,
we constructed linear mixed effects models, as implemented
in the lme4 package (Bates et al., 2015). The full model
included geographic location, binary management practice,
and soil physical/chemical characteristics. A random effect for
“Sampling Kit” (a proxy for individual gardens) was included
to account for pseudoreplication, that is, samples from differ-
ent beds in the same garden. We used package lmerTest to
conduct stepwise top-down model selection based on Akaike
information criteria (AIC) values, as recommended by Zuur
et al. (2009), which resulted in a final model containing none
of the measured explanatory variables. Therefore, we fit a
linear regression model containing “Sampling Kit” as the
only predictor variable. Residuals were visually assessed for
normal distribution.
We used the same linear mixed effects modeling procedure
to test factors influencing skin alpha diversity, again including
“Sampling Kit” as a random factor to adjust for repeated sam-
pling of individuals. Box–Cox transformation was applied to
skin alpha diversity estimates, as residuals failed to meet the
assumption of normality. The final selected model included
geographic location, sample collection time point, number of
hours since previous soil contact, and use of chlorine swim-
ming pool within 24 h of initial sample collection. The final
model was fit using REML and pvalues obtained using
Type II Wald chi-square tests with the car package (Fox &
Weisberg, 2011).
2.5.3 Beta diversity
Permutational multivariate analysis of variance (PER-
MANOVA; Anderson (2017)), as implemented by the
adonis2 function in the vegan package (Oksanen et al.,
2018), was used to test for effects of geographic location,
management practices, crop type, and soil physical/chemical
parameters on garden soil beta diversity, which mea-
sures compositional differences among samples. We used
Morisita–Horn distance to quantify community dissimilarity
and visualized results with principal coordinate analysis
(PCoA) ordinations. Permutation design specified 9999
MHUIREACH ET AL.7of22
permutations and included individual garden site as a block-
ing factor to account for non-independence of the three soil
samples contributed by each study participant. Thus, the
full model used the same predictor variables as were used
in the linear mixed effect model for soil alpha diversity.
Importantly, we used the “term” option of adonis2, which
partials out variation to each factor in the order it appears in
the model. Individual garden site was also tested in a separate
PERMANOVA model to extract its contribution to variation
in soil microbial community dissimilarity. The assumption of
homogeneous dispersions around group centroids (defined
as spatial median) was tested using betadisper followed
by permutest to obtain Fand pvalues. Similar testing
procedures were used for skin microbiome samples, except
that the explanatory variables used in the PERMANOVA
were geographic location, management practices, sample
collection timepoint, amount of time spent gardening prior
to post-gardening sample collection, number of hours passed
between baseline skin swab and any prior gardening activ-
ities, number of times the participant washed their hands
during the 24-h study, whether antibacterial soap or sanitizers
were used, use of a chlorine swimming pool within 24 h of
baseline skin swab, and presence of indoor pets.
2.5.4 Differentially abundant taxa
Finally, we used a generalized linear model (GLM) based on
the negative binomial distribution, as implemented in DESeq2
(Love et al., 2014) to identify taxa that were enriched or
depauperate among groups. For the comparison between soil
and skin swab samples, we performed this analysis at the
phylum level, whereas the comparison between geographic
locations and management practices within soil samples was
performed at the genus level. Due to the high degree of
interindividual variation, we did not test for differentially
abundant taxa within skin swab samples.
3 RESULTS
3.1 Study overview
3.1.1 Soil physical/chemical attributes
In general, gardens in this study tended to have very high
levels of organic matter and soil nutrients compared with
recommended ranges for the region (Figure 2and Table
S2). None of these attributes were associated with man-
agement practices (Table S3), although soil pH, nitrate,
phosphate, potassium, and magnesium levels were associated
with geographic location (Figure 2).
3.1.2 Microbial community overview
After quality filtering, there were 8,609,271 total bacterial
16S rRNA reads in garden soil samples, representing 47,101
different ASVs across 31 bacterial phyla. In skin micro-
biome samples, there were 6,161,190 total reads, representing
12,432 different ASVs across 32 bacterial phyla. The most
abundant phyla found in garden soil communities included
Firmicutes (44.8%), Proteobacteria (19.1%), Verrucomicro-
bia (9.1%), and Acidobacteria (6.9%). Skin swab samples
were dominated by members of the same four phyla, albeit
in different proportions: Proteobacteria (78.0%), Firmicutes
(18.2%), Verrucomicrobia (2.3%), and Acidobacteria (0.8%).
3.1.3 Alpha diversity
Garden soils had over 28 times greater mean effective
numbers of species based on Shannon entropy than skin
microbiomes, aggregated across the entire study (Table 1;
Mann–Whitney U test: U=1600, P<0.001). We observed a
marked pattern of greater taxonomic richness in soil samples
compared with skin samples, even when library size (i.e., sam-
pling effort) was similar (Figure S1). The large disparities in
effective numbers of species based on observed taxa, Shannon
entropy, and Gini–Simpson index shown in Table 1indicate
a substantial degree of unevenness in the community struc-
ture, which is more pronounced for skin microbiomes than
for garden soil.
3.1.4 Beta diversity
Sample type (soil vs. skin) explained 25.9% of the varia-
tion in community dissimilarity among all samples (Figure 3;
PERMANOVA: P<0.001). We found that phyla Acidobac-
teria, Verrucomicrobia, and Planctomycetes were the most
enriched taxa in soil, whereas Proteobacteria, Actinobacteria,
and Firmicutes were the most enriched taxa on skin (Figure 4).
3.2 Garden soil bacterial community
structure
3.2.1 Composition
Despite their higher mean effective number of species, garden
soil bacterial communities were more similar in composi-
tion (Figure 5) than skin communities. We found that 26
genera were common to over 95% of all garden soil sam-
ples (Figure S2), which we define as the core garden soil
microbiome in this study. The most abundant genera across
8of22 MHUIREACH ET AL.
Management practices
Nonorganic
Organic
Leaflet | Tiles © Esri — Source: USGS, Esri, TANA, DeLorme, and NPS
FIGURE 2 Boxplots for each soil physical and chemical attribute, as a function of garden location. Adjusted pvalues for Welch’s t-tests are
annotated at the top of each plot. pvalues were adjusted using the Benjamini–Hochberg method to control false discovery rate. Raw untransformed
values for individual soil samples are plotted as points overlaid on boxplots. Point color indicates management practices. Recommended ranges from
Sullivan et al. (2011), if available, are indicated with dashed horizontal lines.
TABLE 1 Mean effective numbers of species for garden and skin samples based on observed counts, Shannon entropy, and Gini–Simpson
index. Standard deviations for each mean value are shown in parentheses. Groups are further broken into subgroups by location and management
practices.
Type NObserved Shannon Simpson
Garden 120 3387.7 (±89.9) 1324.5 (±36.5) 477.7 (±18.3)
G_High Desert 60 3515.1 (±139.3) 1359.8 (±54.8) 491.4 (±24.2)
G_Willamette Valley 60 3260.3 (±112.4) 1289.3 (±48.3) 464.1 (±27.5)
G_Nonorganic 60 3333.8 (±129.4) 1290.2 (±52.3) 466.4 (±26.6)
G_Organic 60 3441.6 (±125.5) 1358.9 (±51.0) 489.1 (±25.3)
Skin 155 267.1 (±28.0) 46.3 (±10.9) 15.1 (±3.9)
S_High Desert 78 229.1 (±27.4) 22.4 (±9.0) 7.2 (±2.6)
S_Willamette Valley 77 305.7 (±48.8) 70.5 (±19.6) 23.0 (±7.3)
S_Nonorganic 77 329.1 (±45.3) 58.8 (±17.0) 18.1 (±6.0)
S_Organic 78 206.0 (±31.9) 33.9 (±13.5) 12.0 (±5.0)
MHUIREACH ET AL.9of22
FIGURE 3 Principal coordinates analysis (PCoA) ordination plot showing community dissimilarity (based on Morisita–Horn distance) among
samples. Point color indicates sample type (garden soil, skin), as well as sample collection time for skin swab samples.
all garden soil samples were Sphingomonas (5.3%), RB41
(2.6%), and Candidatus_Udaeobacter (2.5%).
3.2.2 Alpha diversity
Within garden soil samples, none of the a priori variables
of interest, including geographic location, management prac-
tices, and crop type, were associated with alpha diversity after
controlling for the effect of pseudoreplication (i.e., three sam-
ples from each garden; Table S4). However, the effect of
garden site itself was a significant influence on alpha diversity
(Linear Regression Model: F =5.8, P<0.001).
3.2.3 Beta diversity
Overall, the PERMANOVA model explained 43.6% of the
variation in community similarity (Figure 6). After con-
trolling for the effect of pseudoreplication, the strongest
predictive factors were geographic location and crop type;
a number of physical and chemical soil parameters also had
significant, but weak, marginal effects (Table 2). However,
similar to alpha diversity analysis, the effect of individual
garden site, which was used as a blocking factor in the
multivariate PERMANOVA model, was far stronger than
TABLE 2 PERMANOVA results for factors influencing garden
soil bacterial community similarity. Factors are listed in the table in the
same order as in the model formula, thus the reported statistics and p
values represent marginal effects after accounting for factors that
appeared earlier in the model formula.
Variable df Sum of squares R2Fpvalue
Location 1 1.95 0.09 14.38 0.0001
MP.binary 10.31 0.01 2.25 0.0001
CropFamily 14 3.93 0.17 2.07 0.0001
pH 10.77 0.03 5.65 0.0076
OM 1 0.89 0.04 6.57 0.001
NO3.N 10.32 0.01 2.38 0.0008
PO4.P 1 0.24 0.01 1.75 0.2187
K 1 0.42 0.02 3.11 0.0021
Ca 1 0.50 0.02 3.69 0.0012
Mg 10.35 0.02 2.58 0.0283
Location:MP.binary 1 0.27 0.01 2.02 0.647
Residual 95 12.89 0.56 NA NA
Total 119 22.84 1.00 NA NA
any of the variables of interest (R2=0.62, P<0.001).
Management practices were not associated with garden soil
community similarity.
10 of 22 MHUIREACH ET AL.
SkinGarden
Acidobacteria
Gemmatimonadetes
Armatimonadetes
Verrucomicrobia
Planctomycetes
Chlamydiae
Thaumarchaeota
Chloroflexi
Nitrospirae
Hydrogenedentes
Dependentiae
Latescibacteria
Actinobacteria
Fibrobacteres
Spirochaetes
Deinococcus−Thermus
Halanaerobiaeota
Omnitrophicaeota
Patescibacteria
Tenericutes
Cyanobacteria
Coprothermobacteraeota
Atribacteria
Thermotogae
Elusimicrobia
Synergistetes
Lentisphaerae
Epsilonbacteraeota
Proteobacteria
Fusobacteria
Firmicutes
−10 −5 0 5 10
Log2FoldChange
Phylum
FIGURE 4 Phyla that were enriched in either garden soil or skin swab samples.
3.2.4 Differential abundance
Because our a priori hypothesis was that management
practices and geographic location influence garden soil
microbiomes, we further investigated potential discrimination
by individual bacterial taxa using a negative binomial GLM.
We identified relatively few taxa differentiating garden soil
microbial communities by management practices (Figure 7)
compared to a much larger number that were differentially
abundant by location (Figure 8). This reinforces our conclu-
sion that management exerts a weak influence compared with
geographic location.
3.3 Gardeners’ skin bacterial community
structure
3.3.1 Composition
Composition of skin bacterial communities varied dramati-
cally between individuals and across time within the same
individual (Figure 9). In contrast to soil samples, which shared
many taxa, only one genus (Pantoea) was common to >
95% of skin samples. In addition to Pantoea, which com-
prised 16.5% of all skin sample reads, other abundant genera
found on skin included Acinetobacter (29.9%) and Klebsiella
(14.2%).
3.3.2 Alpha diversity
The most influential factors driving skin alpha diversity, after
controlling for pseudoreplication, were geographic location
(Linear Mixed Effects Model: χ2=8.2, P=0.004), sample
collection time point (χ2=19, P<0.001), number of hours
since previous soil contact (χ2=3.6, P=0.059), and use of
a chlorine swimming pool within 24 h of initial sample col-
lection (χ2=7.8, P=0.005). Higher skin alpha diversity
was associated with living in the Willamette Valley, sam-
ples collected immediately after gardening, fewer number of
hours since previous soil contact, and recent use of a chlo-
rine swimming pool. Management practices, however, did not
MHUIREACH ET AL.11 of 22
34 35 36 37 38 39 40 44
25 26 27 28 30 31 32 33
17 18 19 20 21 22 23 24
910 11 12 13 14 15 16
1 2 3 4 5 6 7 8
123 123 123 123 123 123 123 123
123 123 123 123 123 123 123 123
123 123 123 123 123 123 123 123
123 123 123 123 123 123 123 123
123 123 123 123 123 123 123 123
0
25
50
75
100
0
25
50
75
100
0
25
50
75
100
0
25
50
75
100
0
25
50
75
100
Percentage of Sequences
Other
Arenimonas
Candidatus_Udaeobacter
Chthoniobacter
Ellin6055
Nocardioides
Pseudarthrobacter
Pseudomonas
RB41
Sphingomonas
Streptomyces
FIGURE 5 Top ten most abundant bacterial genera found in garden soil samples. All less abundant genera were combined into the group
“Other.” Individual plots (facets) represent unique sampling kits (a proxy for individual gardens); note that sampling kit numbers shown above each
plot are the original assigned numbers and were not altered for participants dropping out of the study or new replacements that were recruited.
Samples collected from each of the three beds within each garden are listed along the xaxis for each facet.
significantly affect skin alpha diversity. Some individuals
(e.g., Sampling Kits 6, 12, and 25) tended to maintain higher
than average alpha diversity across all sample collection
timepoints (Figure S3).
3.3.3 Beta diversity
Skin samples collected from gardeners in the High Desert and
those who self-reported use of organic practices had lower dis-
persion around the group centroid (spatial median) than those
living in the Willamette Valley and reporting use of nonor-
ganic practices (Location: F =11, P=0.001; Management
Practices: F =3.7, P=0.057). Because dispersions around the
group centroids were significantly different and homogeneous
dispersions are a critical assumption of the PERMANOVA
test, we cannot conclude whether centroid locations were sig-
nificantly different. Results of the PERMANOVA test are
provided in Table S5. Similar to soil beta diversity, the indi-
vidual was by far the most influential factor shaping skin
12 of 22 MHUIREACH ET AL.
FIGURE 6 PCoA ordination of garden soil microbial community dissimilarity, using Morisita–Horn distance. Points represent individual
garden soil samples, where color indicates management practices ×location and shape indicates location. Ellipses represent 95% confidence
intervals, assuming a multivariate t-distribution.
bacterial community similarity (R2=0.41, P<0.001). Since
skin microbiomes were highly individualized to each partici-
pant, we did not test for differential abundance by location or
management practices.
3.3.4 Skin microbial community change over
time
Our initial hypothesis was that skin microbiome samples
would be more similar to soil samples immediately after gar-
dening, due to microbial transfer from soil to skin during
direct contact. We also expected that the skin microbiome
would return to baseline (before gardening) after a period
of time, depending on individual behaviors, such as wash-
ing hands and bathing. At least for some individuals, this was
indeed the case. For 27 out of the 33 study participants for
whom we had complete skin swab data, the number of taxa
shared between their skin and garden soil increased imme-
diately after gardening activities (Figure 10), and for these
participants the number of shared taxa increased by 180, on
average. Taxa that were most commonly transferred included
members of Sphingomonadaceae, Nocardioidaceae, Xan-
thobacteraceae, Burkholderiaceae, and Pseudomonadaceae.
However, soil microbes were generally transient on the
skin and were no longer present after 12 h. We note
that our study occurred near the onset of the COVID-19
pandemic in the United States, and that this may have influ-
enced hand-washing behaviors and use of hand sanitizers,
which could have had additional unexpected impacts on the
skin microbiome. Only 1 participant in this study reported
washing their hands fewer than 4 times, 15 participants
reported washing 4–6 times, 15 participants reported washing
7–9 times, and 9 participants reported washing 10 or more
times. Additionally, 14 participants reported using hand san-
itizer and/or antibacterial soap during the study, which is
MHUIREACH ET AL.13 of 22
FIGURE 7 Volcano plot of differentially abundant taxa in pairwise comparison between organic and nonorganic soil samples. For readability,
only significant taxa with the highest foldchanges are labeled.
also likely to have affected skin microbiome diversity and
composition.
4 DISCUSSION
The specific ecological role of most microbes, both in soil
and on skin, is a relatively new area of investigation garner-
ing intense interest. However, few concrete recommendations
are currently available to guide actions toward improving soil
microbiomes for plant and human health. This study reports
preliminary evidence on responses of garden soil microbial
communities to different management practices and location-
specific climatic factors, while also providing baseline data
for future studies.
4.1 Soil physical/chemical attributes
Mirroring other recent studies (e.g., (Nelson et al., 2022;
Salomon et al., 2020;Ugarte&Taylor,2020)), we found
that garden soils in this study tended to have overly enriched
organic matter and other plant macronutrients, in many cases
by orders of magnitude above recommended levels. In partic-
ular, phosphate was substantially higher in Willamette Valley
gardens than High Desert gardens, while the reverse was true
14 of 22 MHUIREACH ET AL.
FIGURE 8 Volcano plot of differentially abundant taxa in pairwise comparison between High Desert and Willamette Valley soil samples. For
readability, only significant taxa with the highest foldchanges are labeled.
for nitrate, potassium, and magnesium. One explanation for
the differing phosphate levels is that High Desert soils tend
to be coarse and sandy, and phosphorus leaches easily from
coarse soils. Willamette Valley soils, in contrast, have high
clay content, which retains phosphorus. Organic matter and
magnesium were also well above recommended levels for gar-
dens in this study (Sullivan et al., 2011), but this enrichment
was not associated with differing management practices or
location. The observed overenrichment of plant macronutri-
ents may reflect the care and intensive management applied
to garden beds, in contrast to underlying parent soils. In
addition, many gardeners in this study reported using raised
beds, which commonly contain large amounts of commer-
cial potting mix, compost, and mulch, possibly resulting in
soil test results that are not indicative of parent soil health
(Nelson et al., 2022).
4.2 Soil bacterial communities were largely
determined by site-specific factors
In general, representatives of phyla Proteobacteria, Aci-
dobacteria, Verrucomicrobia, and Actinobacteria are typically
dominant in urban soil microbiomes (Gómez-Brandón et al.,
2022; Delgado-Baquerizo et al., 2021). However, we observed
an unexpectedly high relative abundance of Firmicutes, which
comprised 44.8% of all reads from garden soil samples. A
recent study comparing soil microbiota inhabiting different
MHUIREACH ET AL.15 of 22
34 35 36 37 38 39 40 44
25 26 27 28 30 31 32 33
17 18 19 20 21 22 23 24
910 11 12 13 14 15 16
1 2 3 4 5 6 7 8
1234 1234 1234 1234 1234 1234 1234 234
1234 1234 1234 1234 1234 124 1234 1234
1234 1234 1234 1234 1234 234 1234 1234
1234 1234 234 1234 1234 1234 1234 1234
1234 124 1234 1234 1234 1234 1234 1234
0
25
50
75
100
0
25
50
75
100
0
25
50
75
100
0
25
50
75
100
0
25
50
75
100
Percentage of Sequences
Other
Acinetobacter
Bacillus
Enhydrobacter
Klebsiella
Paenibacillus
Pantoea
Pseudomonas
Raoultella
Staphylococcus
Stenotrophomonas
FIGURE 9 Top 10 most abundant bacterial genera found in skin swab samples. All less abundant genera were combined into the group
“Other.” Individual plots (facets) represent unique sampling kits (a proxy for individual gardeners); note that the sampling kit numbers shown above
each plot are the original assigned numbers and were not altered for participants dropping out of the study or new replacements that were recruited.
Skin swab samples at each timepoint (1, before gardening; 2, immediately after gardening; 3, 12 h after gardening; 4, 24 h after gardening) are listed
on the xaxis.
urban green space types may help explain this phenomenon—
gardens in that study had very high enrichment of Firmicutes,
contrasting with levels found in sports fields, parklands, and
revegetated areas (Baruch et al., 2021). Long-term compost
amendment typical of residential gardens has been shown
to increase relative abundance of Firmicutes (Kim et al.,
2021), as has shorter-term application of manure (Billet et al.,
2022). Indeed, Firmicutes are the dominant bacterial phyla in
chicken and cattle manures (Zhang et al., 2018;Kim&Wells,
2016) and are also highly abundant in horse manure (Salem
et al., 2018), several commonly used types of organic fertilizer
for gardens.
Compared with skin, garden soils displayed much greater
alpha diversity, in terms of both richness and evenness.
16 of 22 MHUIREACH ET AL.
6
7
8
12
24
25
32
1
3
5
6
7
8
10
12
16
18
19
24
25
28
36 7
8
10
12
25
37
5
6
8
12
16
19
25
0
500
1000
Skin Before Skin After Skin After 12 Hours Skin After 24 Hours
Number of taxa shared between skin and soil
FIGURE 10 Taxa shared between gardeners’ skin and their garden soil increase immediately after gardening but return to baseline by 12 h
afterward. Shared taxa were calculated by presence–absence comparison between each gardener’s skin samples at the indicated timepoint and their
aggregated garden soil microbial communities. Points are labeled by sampling kit.
However, contrary to our a priori hypotheses, we did not
find evidence that geographic location, management prac-
tices, crop type, or soil physical/chemical parameters exerted
significant influence on alpha diversity. Instead, soil micro-
biome diversity appeared largely driven by (unmeasured)
characteristics of individual gardens. This result is unlike
prior studies of garden soil biodiversity, which have found
that soil-protective management practices, such as applying
compost or avoiding tilling, are associated with increased
belowground biodiversity (Tresch et al., 2019). One explana-
tion for our contrasting result may be that several practices,
such as tilling and applying compost, were popular across
both organic and nonorganic gardeners in our study. Another
possibility is that self-reported management practices may
not have aligned with USDA organic standards, as organic
certifications are typically aimed at commercial producers
rather than gardeners. Similarly, we did not find a substantial
association between alpha diversity and soil pH, which has
been noted in previous work (Gómez-Brandón et al., 2022),
possibly because many gardeners strive (and were success-
ful in this study) to maintain pH within a relatively narrow
range that is favorable for common vegetable crops. A similar
pattern was reflected in beta diversity, which was more
strongly related to individual garden site than to measured
variables of interest. After controlling for the effect of site, we
found that community dissimilarity was primarily associated
with crop type and geographic location, which explained 17%
and 9% of the remaining variation, respectively. All soil phys-
ical and chemical parameters, except phosphate, were also
weak predictors of dissimilarity, each explaining less than 5%
of remaining variation.
Other research has shown that, although relatively rich
in diversity, soil microbiomes in urban green spaces, such
as gardens, tend to be more homogenized at a global scale
than the range of ecosystems and climates within which
they are set (Delgado-Baquerizo et al., 2021). Because the
human diet in developed countries contains relatively few
different plant-based food types, gardeners across the globe
tend to grow a limited variety of crops, which also tend to
grow well under certain soil conditions. Thus, garden soil
MHUIREACH ET AL.17 of 22
microbiomes are relatively homogeneous because gardens
themselves are homogeneous, in comparison to their geo-
graphic, climatic, and macrobiome (e.g., plants, animals, and
other large biota) contexts. In particular, members of bacte-
rial genera Streptomyces and Pseudomonas, which were both
relatively abundant in garden soils in this study, have been
identified as potential urban greenspace generalists (Delgado-
Baquerizo et al., 2021). However, when focusing on the
variation that does exist across private gardens, results of
this study and others (e.g., Coller et al. (2019) suggest that
latent variables associated with individual site are primary
determinants of microbial community structure.
4.3 Dramatic variation in skin bacterial
communities
The majority of skin swab samples were overwhelmingly
dominated by only a few different taxa, mostly belong-
ing to phylum Proteobacteria. Members of Firmicutes were
also observed in high abundance, however, other phyla
(i.e., Actinobacteria, Bacteroidetes) that are typically highly
represented in human hand microbiomes (Edmonds-Wilson
et al., 2015) were not abundant, possibly because we used
16S primers that are biased toward soil microbiota and
are less effective at amplifying common skin-associated
bacterial taxa.
Unlike soil, gardeners skin microbial community structure
exhibited wide variation between participants and even within
the same participants at different timepoints. For alpha diver-
sity, this variation was best explained by geographic location,
sample collection time point, number of hours since previ-
ous soil contact, and use of a chlorine swimming pool within
24 h of sample collection; management practices were not a
significant factor. Beta diversity and community composition
were so highly variable that we could not conclude whether
any factors aside from the individual subject were influen-
tial. Other hand microbiome studies have also observed a
high degree of inter- and intrapersonal variation (Flores et al.,
2014; Fierer et al., 2008), and that microbiome community
variability itself can be a distinguishing trait between dif-
ferent individuals (Oh et al., 2016). In support of this, we
found that certain individuals (e.g., Sampling Kits 6, 12, 25)
maintained higher than average alpha diversity throughout the
study. However, Oh et al. (2016) also demonstrated that skin
microbial community membership could be stable for up to 2
years, despite frequent perturbations, which contrasts with our
results, possibly due to differing collection schedules and/or
other methodological differences. In particular, hand wash-
ing was reported as frequent in this study, due to the global
COVID-19 pandemic, which may have impacted skin micro-
biome composition. Some studies have found that length of
time since hand washing can influence bacterial composition,
with families Staphylococcaceae, Streptococcaceae, and Lac-
tobacillaceae being more abundant with shorter time since
washing (Fierer et al., 2008). A number of study partici-
pants also reported use of hand sanitizer and/or antibacterial
soap, which may have contributed to the very low diversity
we observed. This may also have health implications, as at
least one prior study has found that microbiome commu-
nity structure is related to both hand hygiene practices and
the likelihood of harboring pathogenic taxa (Rosenthal et al.,
2014).
Despite the variation in community structure across sam-
ples, we observed substantial transfer of soil microbiota to
gardeners’ skin immediately after gardening activities, sub-
stantiating one of our initial hypotheses. The most commonly
transferred microbial taxa included members of Sphingomon-
adaceae, Nocardioidaceae, Xanthobacteraceae, Burkholderi-
aceae, and Pseudomonadaceae, all of which are typically
associated with soil and plant rhizospheres (Glaeser &
Kämpfer, 2014; Tóth & Borsodi, 2014; Oren, 2014; Coenye,
2014; Roquigny et al., 2017). A recent study found that human
populations with closer relationships to natural environments
tend to have more soil-associated microbial taxa, such as
Rhodobacteraceae, Nocardioidaceae, Bacillaceae, Bradyrhi-
zobiaceae and Rhizobiaceae, on their hands than Westernized
populations (Hospodsky et al., 2014). Although the spe-
cific soil-associated families differ (with the exception of
Nocardioidaceae), possibly due to geographic or climatic dif-
ferences, nonetheless it supports our hypothesis of substantial
soil-to-skin transfer.
4.4 Study limitations
As this was a community science effort, some variability
may be attributed to the sample collection process itself,
despite efforts to ensure a standardized protocol by provid-
ing written and video instructions. However, other large-scale
microbiome-focused community science projects (e.g., Amer-
ican Gut Project, Wild Life of Our Homes) have been very
successful, from both scientific and educational perspectives
(McDonald et al., 2018; Dunn et al., 2013).
Because this study was not conceived as an in-depth inves-
tigation of the causes and consequences of skin or soil
microbiome variation, we did not collect detailed demo-
graphic information, such as age, gender, or ethnicity. It is
possible that the degree of gardeners’ cultural diversity could
be related to crop diversity and, thereby, to garden soil micro-
biome diversity, which could be a fruitful area for future
research. Similarly, we did not exclude volunteers from par-
ticipating on the basis of any underlying health or personal
conditions, other than a requirement that study participants
must be adults and should have direct skin-to-soil contact dur-
ing the gardening activity. Some gardens were also located
in rural areas, which could conceivably influence soil charac-
teristics and the likelihood that gardeners would implement
18 of 22 MHUIREACH ET AL.
raised beds versus in-ground systems. Lastly, we did not
anticipate the global COVID-19 pandemic, nor the result-
ing increase in hand washing and hand sanitizer use, which
may also have affected the temporal trajectory of gardeners’
skin microbiomes.
5CONCLUSION
This study represents one of the very first investigations of
garden soil microbiomes and, to our knowledge, the only one
that explores the ability of soil microbes to transfer and per-
sist on human skin after typical gardening activities. Overall,
we found that garden soils tend to have far greater bacterial
diversity than skin microbiome samples. Bacterial commu-
nity composition was largely similar across different garden
beds, whereas skin microbiome composition varied dramat-
ically. Some soil microbes appeared to transfer onto skin
during direct contact with soil, but they were generally gone
within 12 h, suggesting a low ability to permanently colo-
nize skin. However, a daily gardening routine with repeated
and extended contact with soil likely reinoculates the skin
such that soil microbes are often present, holding potential to
impact health. The results of this study may be of interest not
only to hobby gardeners, but also to market growers and other
agricultural producers, and to society at large.
DATA AVAILABILITY STATEMENT
Raw 16S rRNA sequencing data has been deposited in
the NCBI Sequence Read Archive (SRA) database under
BioProject ID PRJNA849001.
ACKNOWLEDGMENTS
This work was supported by an Education and Workforce
Development Postdoctoral Fellowship (Grant No. 2020-
67034-31786) from the USDA National Institute of Food
and Agriculture, Agriculture and Food Research Initiative.
Any opinions, findings, conclusions, or recommendations
expressed in this publication are those of the author(s) and
do not necessarily reflect the view of the U.S. Department
of Agriculture. Funds for soil analyses were provided by the
Oregon State University College of Agricultural Sciences.
We are grateful to the community scientists who volun-
teered their time and soil/skin microbiomes for this study,
and especially to the two gardeners who served on our Advi-
sory Committee and helped us interpret and present results
in the most effective way. We would also like to thank the
other members of the Advisory Committee—Dr. Bart John-
son, Dr. Jonathan Eisen, and Harper Keeler. Special thanks to
Dr. Krista McGuire, who acted as a co-mentor and provided
invaluable advice for microbial analyses and visualizations.
This study was approved by the University of Oregon
Institutional Review Board on May 5, 2020 (Approval No.
03112020.013), and all participants provided informed con-
sent.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest.
AUTHOR CONTRIBUTIONS
Gwynne Mhuireach: Conceptualization; data curation; for-
mal analysis; funding acquisition; investigation; methodol-
ogy; project administration; visualization; writing–original
draft; writing–review and editing. Kevin G. Van Den
Wymelenberg: Conceptualization; project administration;
resources; supervision; writing–review and editing. Gail A.
Langellotto: Conceptualization; data curation; funding acqui-
sition; investigation; methodology; resources; supervision;
writing–review and editing.
ORCID
Gwynne Á. Mhuireach https://orcid.org/0000-0003-2105-
7930
Gail A. Langellotto https://orcid.org/0000-0001-6705-
2559
REFERENCES
Abrahams, P. (2002). Soils: Their implications to human health. Sci-
ence of The Total Environment,291(1), 1–32. https://doi.org/10.1016/
S0048-9697(01)01102-0
Adewopo, J. B., VanZomeren, C., Bhomia, R. K., Almaraz, M., Bacon,
A. R., Eggleston, E., Judy, J. D., Lewis, R. W., Lusk, M., Miller, B.,
Moorberg, C., Snyder, E. H., & Tiedeman, M. (2014). Top-ranked
priority research questions for soil science in the 21 century. Soil Sci-
ence Society of America Journal,78(2), 337. https://doi.org/10.2136/
sssaj2013.07.0291
Anderson, M. J. (2017). Permutational multivariate analysis of vari-
ance (PERMANOVA).Wiley StatsRef: Statistics Reference Online
(pp. 1–15). American Cancer Society. https://doi.org/10.1002/
9781118445112.stat07841
Araújo, A., Leite, L., Santos, V., & Carneiro, R. (2009). Soil microbial
activity in conventional and organic agricultural systems. Sustainabil-
ity,1(2), 268–276.
Bach, J.-F. (2002). The effect of infections on susceptibility to autoim-
mune and allergic diseases. New England Journal of Medicine,
347(12), 911–920. https://doi.org/10.1056/NEJMra020100
Banerjee, S., Walder, F., Büchi, L., Meyer, M., Held, A. Y., Gattinger,
A., Keller, T., Charles, R., & van der Heijden, M. G. A. (2019). Agri-
cultural intensification reduces microbial network complexity and
the abundance of keystone taxa in roots. The ISME Journal,13(7),
1722–1736.
Baruch, Z., Liddicoat, C., Cando-Dumancela, C., Laws, M., Morelli,
H., Weinstein, P., Young, J. M., & Breed, M. F. (2021). Increased
plant species richness associates with greater soil bacterial diversity
in urban green spaces. Environmental Research,196, 110425.
Bates, D., Mächler, M., Bolker, B., & Walker, S. (2015). Fitting lin-
ear mixed-effects models using lme4. Journal of Statistical Software,
67(1). https://doi.org/10.18637/jss.v067.i01
Billet, L., Pesce, S., Martin-Laurent, F., & Devers-Lamrani, M. (2022).
Experimental evidence for manure-borne bacteria invasion in soil
MHUIREACH ET AL.19 of 22
during a coalescent event: Influence of the antibiotic sulfamethazine.
Microbial Ecology.https://doi.org/10.1007/s00248-022-02020-w
Blaser, M. J. (2014). The microbiome revolution. Journal of Clinical
Investigation,124(10), 4162–4165.
Blaser, M. J., & Falkow, S. (2009). What are the consequences of the dis-
appearing human microbiota? Nature Reviews Microbiology,7(12),
887–894.
Blum, W. E., Zechmeister-Boltenstern, S., & Keiblinger, K. M. (2019).
Does soil contribute to the human gut microbiome? Microorganisms,
7(9), 287.
Brown, M. D., Shinn, L. M., Reeser, G., Browning, M., Schwingel,
A., Khan, N. A., & Holscher, H. D. (2022). Fecal and soil micro-
biota composition of gardening and non-gardening families. Scientific
Reports,12(1), 1595.
Callahan, B. J., McMurdie, P. J., Rosen, M. J., Han, A. W., Johnson,
A. J. A., & Holmes, S. P. (2016). DADA2: High-resolution sample
inference from illumina amplicon data. Nature Methods,13(7), 581–
583.
Caporaso, J. G., Lauber, C. L., Walters, W. A., Berg-Lyons, D., Huntley,
J.,Fierer,N.,Owens,S.M.,Betley,J.,Fraser,L.,Bauer,M.,Gormley,
N., Gilbert, J. A., Smith, G., & Knight, R. (2012). Ultra-high-
throughput microbial community analysis on the illumina HiSeq and
MiSeq platforms. The ISME Journal,6(8), 1621–1624.
Castle, S. C., Samac, D. A., Sadowsky, M. J., Rosen, C. J., Gutknecht,
J. L. M., & Kinkel, L. L. (2019). Impacts of sampling design on
estimates of microbial community diversity and composition in agri-
cultural soils. Microbial Ecology,78, 753–763. https://doi.org/10.
1007/s00248-019-01318-6
Chao, A., Gotelli, N. J., Hsieh, T. C., Sander, E. L., Ma, K. H., Colwell,
R. K., & Ellison, A. M. (2014). Rarefaction and extrapolation with
hill numbers: A framework for sampling and estimation in species
diversity studies. Ecological Monographs,84(1), 45–67.
Chen, H., Dai, Z., Veach, A. M., Zheng, J., Xu, J., & Schadt, C. W.
(2020). Global meta-analyses show that conservation tillage practices
promote soil fungal and bacterial biomass. Agriculture, Ecosystems &
Environment,293, 106841.
Coenye, T. (2014). The family Burkholderiaceae. In E. Rosenberg, E.
F. DeLong, S. Lory, E. Stackebrandt, & F. Thompson (Eds.), The
Prokaryotes (pp. 759–776). Springer. https://doi.org/10.1007/978-3-
642-30197-1_239
Coller, E., Cestaro, A., Zanzotti, R., Bertoldi, D., Pindo, M., Larger, S.,
Albanese, D., Mescalchin, E., & Donati, C. (2019). Microbiome of
vineyard soils is shaped by geography and management. Microbiome,
7(1), 140. https://doi.org/10.1186/s40168-019-0758-7
Davis, N. M., Proctor, D. M., Holmes, S. P., Relman, D. A., & Callahan,
B. J. (2018). Simple statistical identification and removal of contami-
nant sequences in marker-gene and metagenomics data. Microbiome,
6(1), 226. https://doi.org/10.1186/s40168-018-0605-2
Delgado-Baquerizo, M., Eldridge, D. J., Liu, Y.-R., Sokoya, B., Wang,
J.-T., Hu, H.-W., He, J.-Z., Bastida, F., Moreno, J. L., Bamigboye, A.
R., Blanco-Pastor, J. L., Cano-Díaz, C., Illán, J. G., Makhalanyane, T.
P., Siebe, C., Trivedi, P., Zaady, E., Verma, J. P., Wang, L., . .. Fierer,
N. (2021). Global homogenization of the structure and function in
the soil microbiome of urban greenspaces. Science Advances,7(28),
eabg5809.
Deng, C., Daley, T., & Smith, A. (2015). Applications of species accu-
mulation curves in large-scale biological data analysis. Quantitative
Biology,3(3), 135–144.
Doran, J. W. (2002). Soil health and global sustainability: Translating
science into practice. Agriculture, Ecosystems & Environment,88(2),
119–127.
Doran, J. W., & Safley, M. (1997). Defining and assessing soil health and
sustainable productivity. In C. Pankhurst, B. M. Doube, & V. V. S. R.
Gupta (Eds.), Biological indicators of soil health. CAB International.
Dumontet, S., Cavoski, I., Ricciuti, P., Mondelli, D., Jarrar, M., Pasquale,
V., & Crecchio, C. (2017). Metabolic and genetic patterns of soil
microbial communities in response to different amendments under
organic farming system. Geoderma,296, 79–85.
Dunn, R. R., Fierer, N., Henley, J. B., Leff, J. W., & Menninger, H. L.
(2013). Home life: Factors structuring the bacterial diversity found
within and between homes. PLoS ONE,8(5), e64133. https://doi.org/
10.1371/journal.pone.0064133
Edmonds-Wilson, S. L., Nurinova, N. I., Zapka, C. A., Fierer, N., &
Wilson, M. (2015). Review of human hand microbiome research.
Journal of Dermatological Science,80(1), 3–12.
Ege, M. J., Mayer, M., Normand, A.-C., Genuneit, J., Cookson,
W. O., Braun-Fahrländer, C., Heederik, D., Piarroux, R., & von
Mutius, E. (2011). Exposure to environmental microorganisms and
childhood asthma. New England Journal of Medicine,364(8), 701–
709. 00265. http://www.nejm.org/doi/full/10.1056/nejmoa10073
02
Esperschütz, J., Gattinger, A., Mäder, P., Schloter, M., & Fließbach, A.
(2007). Response of soil microbial biomass and community structures
to conventional and organicfarming systems under identical crop rota-
tions: Response of soil microbial biomass and community structures.
FEMS Microbiology Ecology,61(1), 26–37.
Fierer, N., Hamady, M., Lauber, C. L., & Knight, R. (2008). The influ-
ence of sex, handedness, and washing on the diversity of hand surface
bacteria. Proceedings of the National Academy of Sciences,105(46),
17994–17999.
Fierer, N., Wood, S. A., & Bueno de Mesquita, C. P. (2021). How
microbes can, and cannot, be used to assess soil health. Soil Biology
and Biochemistry,153, 108111.
Flores, G. E., Caporaso, J. G., Henley, J. B., Rideout, J. R., Domogala, D.,
Chase, J., Leff, J. W., Vázquez-Baeza, Y., Gonzalez, A., Knight, R.,
Dunn, R. R., & Fierer, N. (2014). Temporal variability is a personal-
ized feature of the human microbiome. Genome Biology,15(12), 531.
https://doi.org/10.1186/s13059-014-0531-y
Fox, J., & Weisberg, S. (2011). An R companion to applied regression.
SAGE Publications.
Glaeser, S. P., & Kämpfer, P. (2014). The family Sphingomonadaceae. In
E. Rosenberg, E. F.DeLong, S. Lory, E. Stackebrandt, & F. Thompson
(Eds.), The Prokaryotes (pp. 641–707). Springer. https://doi.org/10.
1007/978-3-642-30197-1_302
Gómez-Brandón, M., Herbón, C., Probst, M., Fornasier, F., Barral, M.,
& Paradelo, R. (2022). Influence of land use on the microbiological
properties of urban soils. Applied Soil Ecology,175, 104452. https://
doi.org/10.1016/j.apsoil.2022.104452
Gotelli, N. J., & Colwell, R. K. (2001). Quantifying biodiversity: Pro-
cedures and pitfalls in the measurement and comparison of species
richness. Ecology Letters,4(4), 379–391.
Grönroos, M., Parajuli, A., Laitinen, O. H., Roslund, M. I., Vari, H. K.,
Hyöty, H., Puhakka, R., & Sinkkonen, A. (2018). Short-term direct
contact with soil and plant materials leads to an immediate increase
in diversity of skin microbiota. MicrobiologyOpen,8(3), e00645.
Guilland, C., Maron, P. A., Damas, O., & Ranjard, L. (2018). Biodi-
versity of urban soils for sustainable cities. Environmental Chemistry
Letters,16(4), 1267–1282.
Haahtela, T. (2019). A biodiversity hypothesis. Allergy,74, 1445–1456.
Hallett, S., Hoagland, L., & Toner, E. (2016). Urban agriculture: Envi-
ronmental, economic, and social perspectives, (pp. 65–120). John
Wiley & Sons, Inc. https://doi.org/10.1002/9781119281269.ch2
20 of 22 MHUIREACH ET AL.
Hanski, I., von Hertzen, L., Fyhrquist, N., Koskinen, K., Torppa, K.,
Laatikainen, T., Karisola, P., Auvinen, P., Paulin, L., Makela, M. J.,
Vartiainen, E., Kosunen, T. U., Alenius, H., & Haahtela, T. (2012).
Environmental biodiversity, human microbiota, and allergy are inter-
related. Proceedings of the National Academy of Sciences,109(21),
8334–8339. 00065. https://doi.org/10.1073/pnas.1205624109
Hartmann, M., Frey, B., Mayer, J., Mäder, P., & Widmer, F. (2015).
Distinct soil microbial diversity under long-term organic and conven-
tional farming. The ISME Journal,9(5), 1177–1194.
Henneron, L., Bernard, L., Hedde, M., Pelosi, C., Villenave, C., Chenu,
C., Bertrand, M., Girardin, C., & Blanchart, E. (2015). Fourteen
years of evidence for positive effects of conservation agriculture and
organic farming on soil life. Agronomy for Sustainable Development,
35(1), 169–181.
Hirt, H. (2020). Healthy soils for healthy plants for healthy humans:
How beneficial microbes in the soil, food and gut are interconnected
and how agriculture can contribute to human health. EMBO Reports,
21(8), e51069. https://doi.org/10.15252/embr.202051069
Hospodsky, D., Pickering, A. J., Julian, T. R., Miller, D., Gorthala, S.,
Boehm, A. B., & Peccia, J. (2014). Hand bacterial communities vary
across two different human populations. Microbiology,160(6), 1144–
1152.
Hsieh, T. C., Ma, K. H., & Chao, A. (2016). iNEXT: An r package for rar-
efaction and extrapolation of species diversity (hill numbers). Meth-
ods in Ecology and Evolution,7(12), 1451–1456. https://besjournals.
onlinelibrary.wiley.com/doi/pdf/10.1111/2041-210X.12613
Ishaq, S. L., Rapp, M., Byerly, R., McClellan, L. S., O’Boyle, M. R.,
Nykanen, A., Fuller, P. J., Aas, C., Stone, J. M., Killpatrick, S.,
Uptegrove, M. M., Vischer, A., Wolf, H., Smallman, F., Eymann, H.,
Narode, S., Stapleton, E., Cioffi, C. C., & Tavalire, H. F. (2019). Fram-
ing the discussion of microorganisms as a facet of social equity in
human health. PLOS Biology,17(11), e3000536. https://doi.org/10.
1371/journal.pbio.3000536
Jost, L. (2007). Partitioning diversity into independent alpha and beta
components. Ecology,88(10), 2427–2439.
Kibblewhite, M., Ritz, K., & Swift, M. (2008). Soil health in agricul-
tural systems. Philosophical Transactions of the Royal Society B:
Biological Sciences,363(1492), 685–701.
Kim, M., & Wells, J. E. (2016). A meta-analysis of bacterial diversity in
the feces of cattle. Current Microbiology,72(2), 145–151.
Kim, S., Samaddar, S., Chatterjee, P., Choudhury, A. R., Choi, J., Choi,
J., & Sa, T. (2021). Structural and functional shift in soil bacte-
rial community in response to long-term compost amendment in
paddy field. Applied Sciences,11(5), 2183. https://doi.org/10.3390/
app11052183
Kong, H. H., Andersson, B., Clavel, T., Common, J. E., Jackson, S. A.,
Olson, N. D., Segre, J. A., & Traidl-Hoffmann, C. (2017). Perform-
ing skin microbiome research: A method to the madness. Journal of
Investigative Dermatology,137(3), 561–568.
Kuras, E. R., Warren, P. S., Zinda, J. A., Aronson, M. F., Cilliers, S.,
Goddard, M. A., Nilon, C. H., & Winkler, R. (2020). Urban socioe-
conomic inequality and biodiversity often converge, but not always: A
global meta-analysis. Landscape and Urban Planning,198, 103799.
https://doi.org/10.1016/j.landurbplan.2020.103799
Lehmann, J., Bossio, D. A., Kögel-Knabner, I., & Rillig, M. C. (2020).
The concept and future prospects of soil health. MicrobiologyOpen,
1(10), 544–553. https://doi.org/10.1002/mbo3.762
Lei, S., Xu, X., Cheng, Z., Xiong, J., Ma, R., Zhang, L., Yang, X., Zhu, Y.,
Zhang, B., & Tian, B. (2019). Analysis of the community composition
and bacterial diversity of the rhizosphere microbiome across different
plant taxa. MicrobiologyOpen,8(6), e00762. https://doi.org/10.1002/
mbo3.762
Leong, M., Dunn, R. R., & Trautwein, M. D. (2018). Biodiversity and
socioeconomics in the city: A review of the luxury effect. Biology
Letters,14(5), 20180082. https://doi.org/10.1098/rsbl.2018.0082
Lori, M., Symnaczik, S., Mäder, P., De Deyn, G., & Gattinger, A. (2017).
Organic farming enhances soil microbial abundance and activity—
A meta-analysis and meta-regression. PLOS ONE,12(7), e0180442.
https://doi.org/10.1371/journal.pone.0180442
Love, M. I., Huber, W., & Anders, S. (2014). Moderated estimation of
fold change and dispersion for RNA-seq data with DESeq2. Genome
Biology,15(12), 1–21. 00667.
Lupatini, M., Korthals, G. W., de Hollander, M., Janssens, T. K.
S., & Kuramae, E. E. (2017). Soil microbiome is more heteroge-
neous in organic than in conventional farming system. Frontiers
in Microbiology,7, 2064. https://doi.org/10.3389/fmicb.2016.020
64
Martìnez-Garcìa, L. B., Korthals, G., Brussaard, L., Jørgensen, H. B.,
& De Deyn, G. B. (2018). Organic management and cover crop
species steer soil microbial community structure and functionality
along with soil organic matter properties. Agricuture, Ecosystems &
Environment,263, 7–17.
Matthews, D. M., & Jenks, S. M. (2013). Ingestion of mycobacterium
vaccae decreases anxiety-related behavior and improves learning in
mice. Behavioural Processes,96, 27–35.
McClintock, N., Young, J. R., Evans, T., Simpson, M., & Santos, J.
(2013). The landscape of food production. Metroscape,8, 13–19.
McDade, T. W. (2012). Early environments and the ecology of inflam-
mation. Proceedings of the National Academy of Sciences,109,
17281–17288.
McDonald, D., Hyde, E., Debelius, J. W., Morton, J. T., Gonzalez, A.,
Ackermann, G., Aksenov, A. A., Behsaz, B., Brennan, C., Chen, Y.,
Goldasich, L. D., Dorrestein, P. C., Dunn, R. R., Fahimipour, A. K.,
Gaffney, J., Gilbert, J. A., Gogul, G., Green, J. L., Hugenholtz, P.,
.. . Knight, R. American gut: An open platform for citizen science
microbiome research. mSystems,3(3), 28.
Ngai,M.,Hadfield,M.G.,Bosch,T.C.G.,Carey,H.V.,Domazet-
Lošo, T., Douglas, A. E., Dubilier, N., Eberl, G., Fukami, T., Gilbert,
S. F., Hentschel, U., King, N., Kjelleberg, S., Knoll, A. H., Kremer,
N., Mazmanian, S. K., Metcalf, J. L., Nealson, K., Pierce, N. E., .. .
Wernegreen, J. J. (2013). Animals in a bacterial world, a new imper-
ative for the life sciences. Proceedings of the National Academy of
Sciences,110(9), 3229–3236.
McGinley, K. J., Larson, E. L., & Leyden, J. J. (1988). Composition and
density of microflora in the subungual space of the hand. Journal of
Clinical Microbiology,26(5), 950–953.
McMurdie, P. J., & Holmes, S. (2013). phyloseq: An r package for
reproducible interactive analysis and graphics of microbiome cen-
sus data. PLOS ONE,8(4), e61217. 00215. https://doi.org/10.1371/
journal.pone.0061217
McPhearson, T., Hamstead, Z. A., & Kremer, P. (2014). Urban ecosystem
services for resilience planning and management in new york city.
AMBIO,43(4), 502–515.
Mills, J. G., Weinstein, P., Gellie, N. J. C., Weyrich, L. S., Lowe, A. J.,
& Breed, M. F. (2017). Urban habitat restoration provides a human
health benefit through microbiome rewilding: The microbiome
rewilding hypothesis. Restoration Ecology,25(6), 866–872.
Moebius-Clune, B., Moebius-Clune, D., Gugino, B., Idowu, O.,
Schindelbeck, R., Ristow, A., van Es, H., Thies, J., Shayler, H.,
McBride, M., Kurtz, K., Wolfe, D., & Abawi, G. (2016). Compre-
MHUIREACH ET AL.21 of 22
hensive assessment of soil health: The Cornell framework manual (3
ed.). Cornell University.
Mulder, I. E., Schmidt, B., Stokes, C. R., Lewis, M., Bailey, M., Aminov,
R. I., Prosser, J. I., Gill, B. P., Pluske, J. R., Mayer, C.-D., Musk,
C. C., & Kelly, D. (2009). Environmentally-acquired bacteria influ-
ence microbial diversity and natural innate immune responses at gut
surfaces. BMC Biology,7, 79. https://doi.org/10.1186/1741-7007-7-
79
Naylor, R. L. (1996). Energy and resource constraints on intensive agri-
cultural production. Annual Review of Energy and the Environment,
21(1), 99–123.
Nelson, M., Mhuireach, G., & Langellotto, G. A. (2022). Excess fertil-
ity in residential-scale urban agriculture soils in two western oregon
cities, USA. Urban Agriculture & Regional Food Systems,7(1),
20027. https://doi.org/10.1002/uar2.20027
Norris, C. E., Bean, G. M., Cappellazzi, S. B., Cope, M., Greub,
K. L., Liptzin, D., Rieke, E. L., Tracy, P. W., Morgan, C. L., &
Honeycutt, C. W. (2020). Introducing the North American project to
evaluate soil health measurements. Agronomy Journal,112(4), 3195–
3215.
Nurminen, N., Lin, J., Grönroos, M., Puhakka, R., Kramna, L., Vari,
H. K., Viskari, H., Oikarinen, S., Roslund, M., Parajuli, A., Tyni, I.,
Cinek, O., Laitinen, O., Hyöty, H., & Sinkkonen, A. (2018). Nature-
derived microbiota exposure as a novel immunomodulatory approach.
Future Microbiology,13(7), 737–744.
Oh, J., Byrd, A. L., Park, M., & Kong, H. H. (2016). Temporal stability
of the human skin microbiome. Cell,165(4), 854–866.
Oksanen, J., Blanchet, F. G., Friendly, M., Kindt, R., Legendre, P.,
McGlinn, D., Minchin, P. R., O’Hara, R. B., Simpson, G. L., Solymos,
P., Stevens, M. H. H., Szoecs, E., & Wagner, H. (2018). Vegan:
Community ecology package (Version R package version 894 2.5-1).
https://CRAN.R-project.org/package=vegan
Oliver, M. A. (1997). Soil and human health: A review. European Jour-
nal of Soil Science,48(4), 573–592. https://onlinelibrary.wiley.com/
doi/pdf/10.1111/j.1365-2389.1997.tb00558.x
Oren, A. (2014). The family Xanthobacteraceae. In E. Rosenberg, E.
F. DeLong, S. Lory, E. Stackebrandt, & F. Thompson (Eds.), The
Prokaryotes (pp. 709–726). Springer. https://doi.org/10.1007/978-3-
642-30197-1_258
O’Riordan, R., Davies, J., Stevens, C., Quinton, J. N., & Boyko, C.
(2021). The ecosystem services of urban soils: A review. Geoderma,
395, 115076. https://doi.org/10.1016/j.geoderma.2021.115076
Pouyat, R. V., Szlavecz, K., Yesilonis, I. D., Groffman, P. M., & Schwarz,
K. (2010). Chemical, physical, and biological characteristics of urban
soils. In J. Aitkenhead-Peterson & A. Volder (Eds.), Urban ecosystem
ecology, number 55 in Agronomy Monograph (pp. 119–152). Amer-
ican Society of Agronomy, Crop Science Society of America, Soil
Science Society of America.
R Development Core Team. (2010). R: A language and environment for
statistical computing.http://cran.r-project.org
Rook, G., Bäckhed, F., Levin, B. R., McFall-Ngai, M. J., & McLean,
A. R. (2017). Evolution, human-microbe interactions, and life history
plasticity. Lancet,390(10093), 521–530.
Rook, G. A. (2013). Regulation of the immune system by biodiver-
sity from the natural environment: An ecosystem service essential to
health. Proceedings of the National Academy of Sciences of the United
States of America,110(46), 18360–18367.
Rook, G. A. W. (2021). Darwinian medicine: We evolved to require
continuing contact with the microbiota of the natural environment.
Evolution turns the inevitable into a necessity. In C. J. Hurst (Ed.
), Microbes: The foundation stone of the biosphere,Advances in
environmental microbiology (pp. 327–364). Springer International
Publishing. https://doi.org/10.1007/978-3-030-63512-1_18
Rook, G. A. W., Adams, V., Palmer, R., Brunet, L. R., Hunt, J., &
Martinelli, R. (2003). Mycobacteria and other environmental organ-
isms as immunomodulators for immunoregulatory disorders. Springer
Seminars in Immunopathology,25(3), 237–255.
Rook, G. A. W., Raison, C. L., & Lowry, C. A. (2014). Microbial “old
friends”, immunoregulation and socioeconomic status. Clinical &
Experimental Immunology,177(1), 1–12.
Roquigny, R., Novinscak, A., Biessy, A., & Filion, M. (2017). Pseu-
domonadaceae: From biocontrol to plant growth promotion. In S.
Mehnaz (Ed. ), Rhizotrophs: Plant growth promotion to bioreme-
diation (pp. 39–68). Springer. https://doi.org/10.1007/978-981-10-
4862-3_3
Rosenthal, M., Aiello, A., Larson, E., Chenoweth, C., & Foxman, B.
(2014). Healthcare workers’ hand microbiome may mediate carriage
of hospital pathogens. Pathogens,3(1), 1–13.
Roslund, M. I., Puhakka, R., Grönroos, M., Nurminen, N., Oikarinen, S.,
Gazali, A. M., Cinek, O., Kramná, L., Siter, N., Vari, H. K., Soininen,
L., Parajuli, A., Rajaniemi, J., Kinnunen, T., Laitinen, O. H., Hyöty,
H., Sinkkonen, A., & ADELE research group. (2020). Biodiversity
intervention enhances immune regulation and health-associated com-
mensal microbiota among daycare children. Science Advances,6(42),
eaba2578.
Ruokolainen, L., von Hertzen, L., Fyhrquist, N., Laatikainen, T.,
Lehtomäki, J., Auvinen, P., Karvonen, A. M., Hyvärinen, A.,
Tillmann, V., Niemelä, O., Knip, M., Haahtela, T., Pekkanen, J.,
& Hanski, I. (2015). Green areas around homes reduce atopic
sensitization in children. Allergy,70(2), 195–202.
Salem, S. E., Maddox, T. W., Berg, A., Antczak, P., Ketley, J. M.,
Williams, N. J., & Archer, D. C. (2018). Variation in faecal micro-
biota in a group of horses managed at pasture over a 12-month period.
Scientific Reports,8(1), 8510. https://doi.org/10.1038/s41598-018-
26930-3
Salomon, M., Watts-Williams, S., McLaughlin, M., & Cavagnaro,
T. (2020). Urban soil health: A city-wide survey of chemical
and biological properties of urban agriculture soils. Journal of
Cleaner Production,275, 122900. https://doi.org/10.1016/j.jclepro.
2020.122900
Salomon, M. J., & Cavagnaro, T. R. (2022). Healthy soils: The back-
bone of productive, safe and sustainable urban agriculture. Journal of
Cleaner Production, 130808. https://doi.org/10.1016/j.jclepro.2022.
130808
Schmid, M. W., Hahl, T., van Moorsel, S. J., Wagg, C., De Deyn, G. B.,
& Schmid, B. (2019). Feedbacks of plant identity and diversity on the
diversity and community composition of rhizosphere microbiomes
from a long-term biodiversity experiment. Molecular Ecology,28(4),
863–878.
Schmidt, B., Mulder, I. E., Musk, C. C., Aminov, R. I., Lewis, M.,
Stokes, C. R., Bailey, M., Prosser, J. I., Gill, B. P., Pluske, J. R., &
Kelly, D. (2011). Establishment of normal gut microbiota is compro-
mised under excessive hygiene conditions. PLoS ONE,6(12), e28284.
https://doi.org/10.1371/journal.pone.0028284
Schnorr, S. L. (2020). The soil in our microbial DNA informs about
environmental interfaces across host and subsistence modalities.
Philosophical Transactions of the Royal Society B,375, 13. https://
doi.org/10.1098/rstb.2019.0577
Schram-Bijkerk, D., Otte, P., Dirven, L., & Breure, A. M. (2018). Indica-
tors to support healthy urban gardening in urban management. Science
22 of 22 MHUIREACH ET AL.
of The Total Environment,621, 863–871. https://doi.org/10.1016/j.
scitotenv.2017.11.160
Stein, M. M., Hrusch, C. L., Gozdz, J., Igartua, C., Pivniouk, V., Murray,
S. E., Ledford, J. G., Marques dos Santos, M., Anderson, R. L.,
Metwali, N., Neilson, J. W., Maier, R. M., Gilbert, J. A., Holbreich,
M., Thorne, P. S., Martinez, F. D., von Mutius, E., Vercelli, D., Ober,
C., & Sperling, A. I. (2016). Innate immunity and asthma risk in
amish and hutterite farm children. New England Journal of Medicine,
375(5), 411–421.
Sullivan, D. M., Horneck, D. A., Owen, J., & Hart, J. M. (2011). Soil test
interpretation guide (EC 1478). Oregon State University Extension
Service.
Taylor, J. R., & Lovell, S. T. (2014). Urban home food gardens in the
global north: Research traditions and future directions. Agriculture
and Human Values,31(2), 285–305.
Taylor, J. R., & Lovell, S. T. (2015). Urban home gardens in the global
north: A mixed methods study of ethnic and migrant home gardens
in chicago, IL. Renewable Agriculture and Food Systems,30(1), 22–
32.
Teuber, S. (2021). A cultural ecosystem service perspective on the inter-
actions between humans and soils in gardens. People and Nature,3(5),
11. https://doi.org/10.1002/pan3.10255
Thompson, L. R., Sanders, J. G., McDonald, D., Amir, A., Ladau, J.,
Locey, K. J., Prill, R. J., Tripathi, A., Gibbons, S. M., Ackermann, G.,
Navas-Molina, J. A., Janssen, S., Kopylova, E., Vázquez-Baeza, Y.,
González, A., Morton, J. T., Mirarab, S., Xu, Z., Jiang, L., .. . Knight,
R. (2017). A communal catalogue reveals earth’s multiscale microbial
diversity. Nature,551(7681), 457–463.
Tiessen, H., Bettany, J., & Stewart, J. (1981). An improved method for
the determination of carbon in soils and soil extracts by dry com-
bustion. Communications in Soil Science and Plant Analysis,12(3),
211–218. https://doi.org/10.1080/00103628109367142
Tilman, D., Cassman, K. G., Matson, P. A., Naylor, R., & Polasky, S.
(2002). Agricultural sustainability and intensive production practices.
Nature,418(6898), 671–677.
Torsvik, V., & Øvreås, L. (2002). Microbial diversity and function in
soil: From genes to ecosystems. Current Opinion in Microbiology,
5(3), 240–245.
Tresch, S., Frey, D., Bayon, R.-C. L., Mäder, P., Stehle, B., Fliessbach,
A., & Moretti, M. (2019). Direct and indirect effects of urban gar-
dening on aboveground and belowground diversity influencing soil
multifunctionality. Scientific Reports,9(1), 9769. https://doi.org/10.
1038/s41598-019-46024-y
Tresch, S., Moretti, M., Le Bayon, R.-C., Mäder, P., Zanetta, A., Frey,
D., & Fliessbach, A. (2018). A gardener’s influence on urban soil
quality. Frontiers in Environmental Science,6, 25. https://doi.org/10.
3389/fenvs.2018.00025
Tóth, E. M., & Borsodi, A. K. (2014). The family Nocardioidaceae. In E.
Rosenberg, E. F. DeLong, S. Lory, E. Stackebrandt, & F. Thompson
(Eds.), The Prokaryotes (pp. 651–694). Springer. https://doi.org/10.
1007/978-3-642-30138-4_193
Ugarte, C. M., & Taylor, J. R. (2020). Chemical and biological indicators
of soil health in chicago urban gardens and farms. Urban Agriculture
& Regional Food Systems,5(1), e20004. https://doi.org/10.1002/uar2.
20004
United Nations. (2019). World urbanization prospects: The 2018 revi-
sion. United Nations, Department of Economic and Social Affairs,
Population Division.
van der Bom, F., Nunes, I., Raymond, N. S., Hansen, V., Bonnichsen,
L., Magid, J., Nybroe, O., & Jensen, L. S. (2018). Long-term fer-
tilisation form, level and duration affect the diversity, structure and
functioning of soil microbial communities in the field. Soil Biology
and Biochemistry,122, 91–103.
Verbruggen, E., Röling, W. F. M., Gamper, H. A., Kowalchuk, G. A.,
Verhoef, H. A., & van der Heijden, M. G. A. (2010). Positive effects
of organic farming on below-ground mutualists: large-scale compar-
ison of mycorrhizal fungal communities in agricultural soils. New
Phytologist,186(4), 968–979.
von Hertzen, L., & Haahtela, T. (2006). Disconnection of man and the
soil: Reason for the asthma and atopy epidemic? Journal of Allergy
and Clinical Immunology,117(2), 334–344. 00059.
von Hertzen, L., Hanski, I., & Haahtela, T. (2011). Natural immunity:
Biodiversity loss and inflammatory diseases are two global mega-
trends that might be related. EMBO Reports,12(11), 1089–1093.
https://doi.org/10.1038/embor.2011.195
Wall, D. H., Nielsen, U. N., & Six, J. (2015). Soil biodiversity and human
health. Nature,528(7580), 69–76.
Wang, Q., Garrity, G. M., Tiedje, J. M., & Cole, J. R. (2007). Naive
Bayesian classifier for rapid assignment of rRNA sequences into the
new bacterial taxonomy. Applied and Environmental Microbiology,
73(16), 5261–5267.
Wickham, H. (2016). Ggplot2: Elegant graphics for data analysis.
Springer-Verlag. https://ggplot2.tidyverse.org
Widmer, F., Rasche, F., Hartmann, M., & Fliessbach, A. (2006). Com-
munity structures and substrate utilization of bacteria in soils from
organic and conventional far ming systems of the DOK long-ter m field
experiment. Applied Soil Ecology,33(3), 294–307.
Zandbergen, P. A. (2014). Ensuring confidentiality of geocoded health
data: Assessing geographic masking strategies for individual-level
data. Advances in Medicine,2014, 1–14. https://doi.org/10.1155/
2014/567049
Zhang, L., Li, L., Pan, X., Shi, Z., Feng, X., Gong, B., Li, J., & Wang, L.
(2018). Enhanced growth and activities of the dominant functional
microbiota of chicken manure composts in the presence of maize
straw. Frontiers in Microbiology,9, 1131. https://doi.org/10.3389/
fmicb.2018.01131
Zuur, A., Ieno, E. N., Walker, N., Saveliev, A. A., & Smith, G. M. (2009).
Mixed effects models and extensions in ecology with R.Springer
Science & Business Media.
SUPPORTING INFORMATION
Additional supporting information can be found online in the
Supporting Information section at the end of this article.
How to cite this article: Mhuireach, G. Á., Van Den
Wymelenberg, K. G., & Langellotto, G. A. (2023).
Garden soil bacteria transiently colonize gardeners’
skin after direct soil contact. Science,8, e20035.
https://doi.org/10.1002/uar2.20035