Skip to main content
Dryad logo

Metadata of literature survey of common gardens


Johnson, Loretta (2021), Metadata of literature survey of common gardens, Dryad, Dataset,


1. Local adaptation is a fundamental phenomenon in evolutionary biology, with relevance for formation of ecotypes, and ultimately new species, and application to restoration and species’ response to climate change. Reciprocal transplants, a common garden in which ecotypes are planted among home and away habitats, is the gold standard to detect local adaptation in populations.

2. This review focuses on reciprocal gardens to detect local adaptation, especially in grassland species beginning with early seminal studies of grass ecotypes. Fast forward more than half a century, reciprocal gardens have moved into the genomic era, in which the genetic underpinnings of ecotypic variation can now be uncovered. Opportunities to combine genomic and reciprocal garden approaches offer great potential to shed light on genetic and environmental control of phenotypic variation. Our decadal study of adaptation in a dominant grass across the precipitation gradient of the US Great Plains combined genomic approaches and realistic community settings to shed light on controls over phenotype.

3. Reciprocal gardens are not without limitations and challenges. A survey of recent studies indicated the modal study uses a tree species, three source sites and one growing site, focuses on one species growing in a monoculture, lasts 3 years, and does not use other experimental manipulations and rarely employs population genetic tools. Reciprocal gardens offer powerful windows into local adaptation, when 1) placed across wide environmental gradients to encompass the species’ range; 2) conducted across timespans adequate for detecting responses; 3) employing selection studies among competing ecotypes in community settings and 4) combined with measurements of form and function which ultimately determine success in home and away environments.

4. Reciprocal gardens have been one of the foundations in evolutionary biology for the study of adaptation for the last century, and even longer in Europe. Moving forward, reciprocal gardens of foundational non-model species, combined with genomic analyses and incorporation of biotic factors, have the potential to further revolutionize evolutionary biology. These field experiments will help us to predict and model response to climate change and to inform restoration practices.


To better understand current trends in studies employing common gardens, we conducted a survey of the literature using reciprocal transplants with plants grown outside published in the last two decades (1990-2020; details are described in Supplemental Methods). In total, 111 relevant articles were assessed. (Spreadsheet of papers assessed are found in supplemental methods.

Literature survey details from supplemental methods

We surveyed the literature to assess the characteristics of studies employing common gardens. A candidate set of articles was obtained using a search on ISI Web of Knowledge (Web of Science; on January 5, 2021. We Used the “advanced” search with the string “TS=(‘COMMON GARDEN*’ OR ‘RECIPROCAL TRANSPLANT*’ OR ‘PROVENANCE’), then selected articles published from 1990 to 2020 (inclusive) and classified in categories “ecology”, “forestry”, “plant sciences”, “evolutionary biology”, “genetics/heredity”, “biodiversity conservation”, “agronomy”, “biology”, “soil science”, or “horticulture.” The search yielded 7,668 articles.

Article titles, abstracts, authors, and DOIs were downloaded, then sorted randomly.  Articles were then assessed in order by which they were sorted. We first assessed whether an article was potentially relevant by scanning the title and abstract; in some cases we also consulted the article text if the title and abstract were not clear on whether common gardens were used or not. Potentially relevant articles that could not be accessed by one team member were attempted by another team member from a different institution.  We also used general searches on for titles of articles which were could not access through our respective library systems. In all, we surveyed the potential relevancy of 408 of the 7,668 articles before being able to assess a sufficient number to ensure statistical stability in the results (see below).

For those articles we assessed as potentially relevant and which we could access and which were written in English, we then assessed whether they 1) used living plants which were 2) grown outside (versus in a growth chamber, greenhouse, shadehouse, etc.). (Articles that used indoor growing arrangements were included if they also used an outdoor growing setting as well.) If an article met these criteria, we assessed a further set of aspects of each study. The number of articles that met these criteria was 111 which was sufficient to ensure statistical reliability (again, see below).

Each relevant article was assessed for (acceptable values are in square brackets):

  • Number of focal species that were plants. A “focal” species was defined as one on which measurements were taken. [integer/unknown]
  • Life form(s) of plants: “tree/shrub” [yes/no], “forb” [yes/no], “grass” [yes/no], or “other” [no/description].
  • Continent(s) of the sites in which plants were grown. [North America/South America/Oceania/Australia/Asia/Africa/Europe/Antarctica/unknown]

The biome(s) of the site in which plants were grown. Plants grown in botanical gardens or other facilities (so long as they were outside) were assigned to the biome in which the growing site was located. In cases where the biome was not described, we inferred the biome by reference to

  •  [alpine/boreal forest/coastal wetland/desert/freshwater wetland/tropical or subtropical grassland/Mediterranean/marine/montane/savanna or scrub woodlands/subtropical and tropical wet or dry forest/temperate forest/unknown]
  • Number of sites from which propagules or plants were sourced. [integer/unknown]
  • Number of sites at which propagules or plants were planted. [integer/unknown]
  • Duration of the study in months, starting when plants were placed/sown outside and ending when plants were measured [integer/unknown]
  • Whether or not focal species were grown alone or in a community with other plant species. To be grown in a community a plant would have reasonably be expected to interact with another species through, for example, root-root or shoot-shoot competition. Articles describing plants being grown in pots or in spaced rows (usually trees) were assumed to be growing plants “alone” unless otherwise noted. [alone/community/both/unknown]
  • Types of measurements taken on the focal species:
    • Mrphology (e.g., specific leaf area, branching, frost damage). [yes/no]
    • Physilogy/growth/metabolites (e.g., respiration, water conductance, chlorophyll α content, biomass or change in biomass, production of primary or secondary compounds). [yes/no]
    • Demgraphy/survivorship/mortality/fecundity (e.g., intrinsic rate of natural increase, percent survival or dead, mass of fruits, number of inflorescences). [yes/no]
    • Phenlogy (e.g., date of budburst, leafout, flowering, fruiting). [yes/no]
    • DNA/RNA/epigenetics (i.e., DNA r RNA extraction, DNA methylation coupled with any analysis). [yes/no]
    • Quantitative genetics (i.e., assessing reactin norms using intentional crosses). [yes/no]
    • Intraspecific cmpetition/facilitation (i.e., any of above as a response to change in conspecific density). [yes/no]
    • Interspecific interactins (i.e., any of above as a response to presence/density of non-plant species including competitors, herbivores, pathogens, facilitators, amensalists, etc.). [yes/no]
  • Number of manipulative treatments (aside from propagation in common gardens). If treatments were applied, we then tallied the types of treatments:
    • Abitic:
      • Soil/nutrients [yes/no]
      • Temperature [yes/no]
      • Water added/removed [yes/no]
      • Altered fire regime [yes/no]
      • Other [no/description]
    • Bitic:
      • Intraspecific competition/facilitation [yes/no]
      • Interspecific competition/facilitation [yes/no]
      • Pollination [yes/no]
      • Herbivory [yes/no]
      • Plant age (e.g., some studies evaluated differences between seedlings and saplings in the same experiment) [yes/no]
      • Introduced/native status (i.e., some studies compared non-native and native ecotypes of the same species in common gardens) [yes/no]
      • Other [no/description]

In a few cases when interpretation was ambiguous, the assessors (LJ, DG, SB, and ABS) consulted with one another. We removed studies that used just one source site and one growing site. We assessed the reliability of inferences from our survey by calculating, for each category scored in a binary manner, the running standard error of the proportion of “yes” responses for categories scores yes/no/unknown. When this was ≤5% for all such categories, we ceased assessment, which occurred at <101 articles (we assessed 111 relevant articles; Fig. S1). The variance for proportional data is equal to p(1 – p), which is maximized when p = 0.5. Thus, the field(s) which were closest to having an equal balance between yes/no responses had the greatest variance (which was Life form: tree/shrub, in our case).  All other fields had a standard error lower than 5%.


U.S. Department of Agriculture, Award: grant 2008-35100-04545

National Science Foundation, Award: GRFP to Galliart