Orchard Management and Landscape Context Mediate the Pear Floral Microbiome

Abstract

Crop-associated microbiota are a key factor affecting host health and productivity. Most crops are grown within heterogeneous landscapes, and interactions between management practices and landscape context often affect plant and animal biodiversity in agroecosystems. However, whether these same factors typically affect crop-associated microbiota is less clear. Here, we assessed whether orchard management strategies and landscape context affected bacterial and fungal communities in pear (Pyrus communis) flowers. We found that bacteria and fungi responded differently to management schemes. Organically certified orchards had higher fungal diversity in flowers than conventional or bio-based integrated pest management (IPM) orchards, but organic orchards had the lowest bacterial diversity. Orchard management scheme also best predicted the distribution of several important bacterial and fungal genera that either cause or suppress disease; organic and bio-based IPM best explained the distributions of bacterial and fungal genera, respectively. Moreover, patterns of bacterial and fungal diversity were affected by interactions between management, landscape context, and climate. When examining the similarity of bacterial and fungal communities across sites, both abundance- and taxon-related turnovers were mediated primarily by orchard management scheme and landscape context and, specifically, the amount of land in cultivation. Our study reveals local- and landscape-level drivers of floral microbiome structure in a major fruit crop, providing insights that can inform microbiome management to promote host health and high-yielding quality fruit. IMPORTANCE Proper crop management during bloom is essential for producing disease-free tree fruit. Tree fruits are often grown in heterogeneous landscapes; however, few studies have assessed whether landscape context and crop management affect the floral microbiome, which plays a critical role in shaping plant health and disease tolerance. Such work is key for identification of tactics and/or contexts where beneficial microbes proliferate and pathogenic microbes are limited. Here, we characterize the floral microbiome of pear crops in Washington State, where major production occurs in intermountain valleys and basins with variable elevation and microclimates. Our results show that both local-level (crop management) and landscape-level (habitat types and climate) factors affect floral microbiota but in disparate ways for each kingdom. More broadly, these findings can potentially inform microbiome management in orchards for promotion of host health and high-quality yields.

Keywords: Pyrus communis; flower microbiome; integrated pest management; landscape heterogeneity.

FIG 1

Relative abundance (proportion of sequences) of bacterial (A) and fungal (B) families associated with pear flowers. Flowers were collected from orchards that reflected three unique management schemes (conventional, bIPM, and organic).

FIG 2

(A) Diversity (Shannon and inverse Simpson indices) of bacteria and fungi associated with pear flowers collected from orchards that vary in management scheme (conventional, bIPM, and organic). (B) Coefficients from the 90% confidence set of top multivariate models. Variable importance was evaluated as the number of models within the 90% confidence model set in which the factor was included.

FIG 3

Canonical correlation analysis of three bacterial and five fungal taxa associated with both disease and disease prevention in pear flowers. The left panels depict the variance explained by the factors in the canonical axes for bacteria (A) and fungi (C), and the right panels depict the variance explained by the canonical axes in the bacterial (B) and fungal (D) taxa of interest.

FIG 4

Restricted distance-based analysis of pear flower bacterial and fungal community beta diversity, including explanatory variables in the top AIC-selected RDA models. Variance explained by each factor can be found in Tables 2 and 3.

FIG 5

Geographic extent of survey, where 15 pear orchards in central Washington across variable landscape contexts were sampled during peak bloom. The study site map was created with ArcGIS (v10.8; ESRI Inc., Redlands, CA, USA) and Inkscape (v1.0.2; https://inkscape.org), with elevation and land cover shading based on the National Elevation Dataset (USDA NRCS) and Cropland Data Layer product (USDA NASS [https://www.nass.usda.gov/Research_and_Science/Cropland/SARS1a.php]), respectively.