TY - JOUR
AU - McGuire,, Krista
AB - ABSTRACT Soil microbes are essential to the continued productivity of sustainably managed agroecosystems. In shade coffee plantations, the relationship between soil microbial composition, soil nutrient availability and coffee productivity have been demonstrated, but the effects of management on the composition of the soil microbial communities remains relatively unexplored. To further understand how management modulates the soil microbiome, the soil fungal and bacterial communities, soil chemistry, and canopy composition were surveyed in a Nicaraguan coffee cooperative, across 19 individual farms. Amplicon sequencing analyses showed that management (organic or conventional), stand age and previous land use affected the soil microbiome, albeit in different ways. Bacterial communities were most strongly associated with soil chemistry, while fungal communities were more strongly associated with the composition of the canopy and historical land use of the coffee plantation. Notably, both fungal and bacterial richness decreased with stand age. In addition to revealing the first in-depth characterization of the soil microbiome in coffee plantations in Nicaragua, these results highlight how fungal and bacterial communities are simultaneously modulated by long-term land use legacies (i.e. an agricultural plot's previous land use) and short-term press disturbance (i.e. farm age). coffee microbiome, sustainable agriculture, fungi, bacteria, microbiome, amplicon sequencing INTRODUCTION The biotic and abiotic drivers of microbial community assembly across space and time determine the composition of a local community, which in turn determines the functional properties of that community (Nemergut et al. 2013). These linkages are especially critical to examine in human-modified terrestrial ecosystems, as microbes perform critical functions related to the resistance and resilience of aboveground plant communities to stress and disturbance (Wardle et al. 2004; van der Putten et al. 2016) and drive the vast majority of biogeochemical cycling in soils (Six et al. 2006; Falkowski, Fenchel and Delong 2008). Agriculture is one of the most widespread and significant ecosystem modifications by humans, and cultivated land now accounts for 36.6% of the terrestrial surface (FAO 2017). Consequently, determining the assembly mechanisms of microbial communities in agroecosystems is essential for leveraging their benefit for crop production, pest management and soil function (Brussaard, De Ruiter and Brown 2007; Berg 2009; Chaparro et al. 2012; Toju et al. 2018). Coffee plantations are important agroecosystems for evaluating the role of microbial assembly mechanisms, as they are one of the most highly traded commodities in the world (Donald 2004; Ricketts et al. 2004) and have significant variation in the intensity by which they are managed (Philpott et al. 2008; Tscharntke et al. 2011; Perfecto, Vandermeer and Philpott 2014). In Central America, coffee production constitutes an essential part of local economies, and coincides with areas of extremely rich biodiversity (Komar 2006). Coffee is especially central to the Nicaraguan economy, where the crop makes up approximately one fifth of the country's exports (Bacon et al. 2008). Following the proliferation of industrial coffee plantations and the extensive clearing of forests for agriculture in Nicaragua in the 1960s–1980s, agrarian reform stimulated the growth of Fair Trade cooperatives (Bacon 2010). International concern about the effects of deforestation on the biodiversity of the region coupled Fair Trade efforts with conservation initiatives, stimulating the creation of shade coffee plantations as faunal reserves (Komar 2006), as well as for the maintenance of ecosystem services (Goodall, Bacon and Mendez 2014; Perfecto, Vandermeer and Philpott 2014). In shade coffee plantations, coffee (Coffea arabica L.) is cultivated, and tree cover is managed to provide shade for the crops. Canopy tree composition may be an important biotic filter for soil microbial communities due to known aboveground–belowground linkages in many terrestrial ecosystems including tropical forests (Peay, Baraloto and Fine 2013; Barberán et al. 2015). Cover crop composition is known to affect microbial composition and function in other agricultural systems (Vukicevich et al. 2018; Hallama et al. 2019; Ma et al. 2020), so it is likely that this also occurs in coffee plantations. Shade cover has been associated to increased coffee market quality (Vaast et al. 2006), as well as higher organic matter content and reduced nutrient leaching (Babbar and Zak 1995; Beer et al. 1997). In practice, canopies in coffee agroecosystems are managed in a highly individualized manner (Hernández-Martínez, Manson and Hernández 2009). The canopy overlying shade coffee plantations may be managed to provide the farmer with additional resources such as timber and fruits. The resulting compositional patchwork likely affects the soil microbiota to different extents. Shade trees in agroecosystems also provide other important ecosystem services, including pollination, pest control and increased connectivity between forest patches in the vicinity of agroecosystems (Philpott et al. 2008; Goodall, Bacon and Mendez 2014; Perfecto, Vandermeer and Philpott 2014). The effect of shade coffee management has assessed for ants (Philpott 2005), birds (Buechley et al. 2015), epiphytes (Goodall, Bacon and Mendez 2014), and woody plants (Kumsa et al. 2016), but few studies have explicitly studied soil microbial communities. To date, studies of soil biota under shade coffee have found that soil carbon associated with microbial biomass is consistently higher in organically managed than in conventional, intensively managed coffee plantations (Partelli et al. 2012), and is associated with higher coffee yields (Silva Aragão et al. 2020). Shade trees can also increase soil N and the abundance of ammonia oxidizing bacteria, which convert ammonia into nitrite in the process of nitrification (Munroe et al. 2015). Together, these studies suggest that management practices, including shade tree management, may directly influence the assembly of soil microbial communities in coffee agroecosystems. The objective of this study was to examine the relationship between management practices, canopy tree composition, soil biochemical composition and soil fungal and bacterial communities in Nicaraguan agricultural cooperatives. We hypothesized that management practices, including whether a farm was conventionally or organically managed, the age of the coffee plantation, and the plot's previous land use, modulate the soil fungal and bacterial communities by altering soil chemistry and aboveground (i.e. canopy) diversity. To this end, 19 coffee farms in Sontule, a small cooperative in the Northern Pacific Nicaraguan highlands, were sampled. The farms varied in time from conversion to coffee plantations, canopy management strategies and coffee plant age, which may affect soil microbial communities due to legacy effects (Kulmatiski and Beard 2008; Fichtner et al. 2014; Schrama et al. 2016). By providing an in-depth examination of how management strategies and the resulting tree composition and soil chemistry influence the soil microbiome, this study informs future shade management strategies. MATERIALS AND METHODS Study site and sample collection The Unión de Cooperativas Agropecuarias (UCA) Miraflor, is a union of small agricultural cooperatives located within the Miraflor Nature Reserve, 30 km from the city of Estelí, Nicaragua (13.25966, −86.243711, Fig. S1, Supporting Information). The UCA consists of approximately 70 farmers who manage several plots of coffee with varying degrees of shade and management intensity. Management strategies were divided into conventional management, which included inputs of herbicide and industrial fertilizer; and organic management, in which soil inputs were limited to various composts. In July 2011, 19 coffee plots were selected after interviewing farmers about their management strategies, age of the plantation, previous land use of the plot, and age of the current coffee plants in order to obtain a representative subset of the UCA (Table S1, Supporting Information). These included plots with a range of coffee plant ages (1:45 years, mean = 12 years) and previous land uses (primary forest, n = 5; secondary forest, n = 7; coffee plantations, n = 7). Five of the plots were conventionally managed, with little or no shade coverage. In addition to the 19 plots, one plot was sampled in unmanaged primary forest, and one was sampled in unmanaged secondary forest. All plots were at an elevation of between 1018 and 1267 m above sea level (mean 1182 m). For each plot, three 5 × 5 m replicate subplots were randomly selected. For each subplot, canopy cover was assessed with a densiometer, measured four times per subplot and averaged. All trees whose canopy fell within the subplot and with diameter breast height greater than 10 cm were identified by their local name and listed. From each subplot, triplicate 5 cm cores were randomly sampled from the top 10 cm soil layer. Triplicate cores from each subplot were later homogenized by sieving through a 2 mm sieve. Soil element analyses were performed at the Auburn University Soil Testing Laboratory (AL, USA), using inductively coupled plasma atomic emission spectroscopy. Soil C and N were measured using loss on ignition. DNA extraction, amplification and sequencing DNA was extracted using the MoBio PowerSoil extraction kit (MoBio, Carlsbad, CA), according to the manufacturer's instructions, with modifications as previously described (Hobbs et al. 2006; Lauber et al. 2009). Bacterial samples were sequenced as part of the Earth Microbiome Project (EMP, Thompson et al. 2017) and fungi were sequenced at the New York Medical College Genomics Core Laboratory (Valhalla, NY). Bacterial communities were assessed by amplifying and sequencing the V3-V4 region of the 16S rRNA gene with primers 515F and 806R (Caporaso et al. 2012), and fungal communities were assessed by amplifying the internal transcribed spacer (ITS1) region with primers ITS1-F and ITS2 (Gardes and Bruns 1993; Bellemain et al. 2010), using adapters developed previously (McGuire et al. 2013). PCR conditions for both 16S and ITS amplification followed protocols and standards detailed in the Earth Microbiome Project (https://www.earthmicrobiome.org/protocols-and-standards/). Both sets of amplicons were sequenced on an Illumina Hiseq with a paired-end 150-bp sequencing kit. Raw sequences were demultiplexed and processed using QIIME (Caporaso et al. 2010) and the UPARSE (Edgar 2013) pipelines. Sequences were quality filtered (maxee = 0.5), with a minimum read length of 150 bp and 250 bp for bacteria and fungi, respectively, and singletons were removed. Sequences were clustered into operational taxonomic units (OTUs) at a 97% similarity threshold and taxonomy was assigned with the RDP classifier (Wang et al. 2007) the Greengenes database (DeSantis et al. 2006) for bacteria, and UNITE for fungi. Datasets were rarefied to 7000 and 20700 sequences per sample for bacteria and fungi, respectively. Only samples for which both fungal bacterial sequences were available were preserved for analyses, resulting in a final dataset which described 19 farms with 48 samples in total (Supporting Information). Bacterial data are publicly available as part of the Earth Microbiome Project (Thompson et al. 2017), and fungal data are publicly available through the Sequence Read Archive, BioProject accession number PRJNA647933. Statistical analyses α-diversity was evaluated as the number of fungal and bacterial OTUs per sample (species richness). Tree richness was measured as the number of different species whose canopy fell within a subplot. Following normality tests, correlations were evaluated with Pearson's R in the case of normal data, and with Spearman's rank correlation for non-normal data. Values are presented as mean ± sd throughout the manuscript. Differences in means were tested non-parametrically with Wilcoxon rank sum tests for comparisons between two groups, and with Kruskal–Wallis tests for comparisons between more groups. Multivariate analyses were performed with the vegan (v.2.5–6) and phyloseq (v.1.26.0) packages (Oksanen et al. 2007; McMurdie and Holmes 2013). Fungal, bacterial, and canopy tree composition as well as soil chemical composition were Hellinger-transformed and converted to Bray–Curtis distances prior to all analyses. The significance of all multivariate analyses was evaluated by permutation tests (adonis, 999 iterations). The relationship between fungal, bacterial, and canopy tree composition, as well as soil chemistry was evaluated in pairwise Mantel tests. The effect of management parameters (Organic/conventional management, the farm's previous land use, and the age of the coffee plants) on fungal and bacterial community composition was evaluated with PERMANOVA-like adonis tests. To investigate the effect of management practices on soil chemistry, a distance-based RDA (db-RDA) was performed on the soil chemistry matrix, constrained by the three management parameters and using Bray-Curtis distances. Interactive terms were not included. The significance of the model, the explanatory terms, and each axis were evaluated by permutations (anova.cca). To further break down the relative contribution of each management parameter, a variance partitioning analysis was performed with the same terms (varpart). Backwards and forwards selection of explanatory variables (ordiR2Step, perm.max = 200) was performed to evaluate the effect of management, canopy composition, and soil chemistry on the soil fungal and bacterial communities. For each community, variables which were selected were included in the db-RDA model, evaluated as previously. Putative functional guilds were assigned to fungal OTUs using the FUNGuild program (Nguyen et al. 2016). To evaluate the effects of management on functional guilds of soil fungi, OTUs assigned as plant pathogens and endophytes were subsetted and agglomerated at the genus level to be analysed separately. Shifts in relative abundances of fungal functional groups across management and age of coffee plants were evaluated using PERMANOVA. A principal component analysis (PCA) was used to construct axes of edaphic variation using the following soil chemical variables: Ca (ppm), K (ppm), Mg (ppm), P (ppm), Al (ppm), Ba (ppm), Fe (ppm), Mn (ppm), Na (ppm), Zn (ppm), B (ppm), Cu (ppm), total %C, total %N and moisture content (%). Relationships between abiotic soil parameters and fungal plant pathogen and endophyte communities were visualized using the first principal component axis and relative abundance data. All data visualization and figures were generated using ‘ggplot2’ (Wickham 2011). RESULTS Average canopy cover did not differ between organically (86.8 ± 6.3%) and conventionally managed farms (84.6 ± 9.7%) or between plots with different historic land uses; however, organically managed farms had a significantly higher canopy tree richness than conventionally managed farms (5.94 ± 1.79 and 3.42 ± 3.02 trees per plot, respectively; Wilcoxon P = 0.001). No relationship between canopy tree richness and historic land use was detected. Soil moisture was highest in plots converted from primary forest, and lowest in plots converted from previous coffee plantations (Kruskall-Wallis χ 2 = 8.52, df = 2, P = 0.014), but no differences between the soil moisture of organically and conventionally managed farms was found. On average across all samples, soil bacterial communities were dominated by Proteobacteria (32.4 ± 3.7% of the community), Verrucomicrobia (17.9 ± 5.4% of the community) and Acidobacteria (16.5 ± 2% of the community). Organically managed plots had consistently higher relative abundances of WS3 (Latescibacteria), while conventionally managed plots had higher relative abundances of Nitrospirae and Chloroflexi (t-tests, P < 0.05 for all comparisons, Fig. 1). Fungal communities were dominated by Ascomycota (37.2 ± 15.5% of the community), Zygomycota (34.6 ± 20.2% of the community) and Basidiomycota (11.9 ± 11.2% of the community). A large portion of the fungal community (12.3 ± 6.7%) was unclassified at the phylum level, and organically managed plots had a higher relative abundance of Ascomycota, while conventionally managed plots had higher relative abundances of Chytridiomycota (Fig. 1). Fungal communities were more variable across plots than bacterial communities (Fig. S2, Supporting Information). Figure 1. Open in new tabDownload slide Relative abundance of the dominant fungal (left) and bacterial (right) phyla on average, for each management type and forest. All fungal phyla are represented. For bacteria, the 10 most dominant phyla on average across all samples are displayed. White asterisks indicate significantly higher relative abundances in a phylum for that management type, as determined by t-tests comparing the conventionally- and organically-managed soils (*** P < 0.001; **P < 0.01; * P < 0.05). The relative abundances of individual samples are available in S2. Figure 1. Open in new tabDownload slide Relative abundance of the dominant fungal (left) and bacterial (right) phyla on average, for each management type and forest. All fungal phyla are represented. For bacteria, the 10 most dominant phyla on average across all samples are displayed. White asterisks indicate significantly higher relative abundances in a phylum for that management type, as determined by t-tests comparing the conventionally- and organically-managed soils (*** P < 0.001; **P < 0.01; * P < 0.05). The relative abundances of individual samples are available in S2. To examine how farm-specific management practices affected soil chemistry, a db-RDA on the soil chemical composition, with organic/conventional management, previous land use, and the age of the coffee plants as explanatory variables was performed. This model explained 27.4% of the variance in soil chemistry and was significant (P = 0.001), as were the three terms added (P ≤ 0.01, Fig. 2). Whether the farm was managed organically or conventionally was the largest driver of soil chemistry (8.2%), followed by the farm's previous land use (7.5%) and the age of the farm's coffee plants (3%). Figure 2. Open in new tabDownload slide A db-RDA shows how different aspects of management affect soil chemistry. Arrows corresponding to conventional management and a previous land use are omitted as they are linearly dependent on the other levels of their respective variables. Circles and triangles represent plots. The db-RDA was significant (P < 0.001), and the significance of each term and axis is indicated with asterisks (*** P < 0.001; **P < 0.01; * P < 0.05). Figure 2. Open in new tabDownload slide A db-RDA shows how different aspects of management affect soil chemistry. Arrows corresponding to conventional management and a previous land use are omitted as they are linearly dependent on the other levels of their respective variables. Circles and triangles represent plots. The db-RDA was significant (P < 0.001), and the significance of each term and axis is indicated with asterisks (*** P < 0.001; **P < 0.01; * P < 0.05). Different aspects of management affected the concentration of specific elements in soil. Organically managed farms had higher percentages of carbon and nitrogen, while conventionally managed farms had higher concentrations of phosphorus (P < 0.05 for all Wilcoxon rank sum tests). In contrast, zinc, aluminium and manganese were significantly higher in farms that were previously coffee plantations, while sodium, carbon, and nitrogen had higher concentrations in lands that were converted from secondary forest, and calcium was highest in lands that were converted from primary forests (P < 0.05 for all Kruskal–Wallis tests). Coffee plant age was negatively correlated with potassium and calcium (R = −0.31 and −0.31, respectively; P < 0.05 for all correlations), and positively correlated with zinc, (R = 0.33, P = 0.02). Alpha diversity Management practices affected the number of bacterial and fungal taxa (richness) in each plot. On average, 680 ± 20 fungal taxa and 1931 ± 83 bacterial taxa were detected per plot. No difference in fungal or bacterial richness between conventionally and organically managed farms was found (P = 0.37 and P = 0.75, respectively). In contrast, the farm's previous land use significantly affected fungal richness (χ2 = 7.94, df = 2, P = 0.019), but not bacterial richness (χ2 = 0.17, df = 2, P = 0.92; Fig. 3A, B). Farms which had been converted from primary forest had lower fungal diversity (575 ± 65 taxa) than those which had been converted from secondary forest or long-term coffee farms (718 ± 113 taxa). Among organically managed farms, both fungal and bacterial richness were negatively and correlated to the age of the coffee plants in the plot, but this was only significant for bacterial richness (R = −0.73, P = 0.007, R2 = 0.53, Fig. 3D). Neither fungal nor bacterial richness in organic farms were significantly correlated with tree canopy richness, however both fungal (R = 0.37, P = 0.03, R2 = 0.14) and bacterial (R = 0.37, P = 0.03, R2 = 0.14) richness among organic farms were positively related to average canopy cover (Fig. 3E and F). Bacterial and fungal richness were weakly correlated to each other (R = 0.28, P = 0.053, R2 = 0.08). Figure 3. Open in new tabDownload slide Soil microbial richness varies according to environmental parameters. Fungal richness is significantly higher in farms converted from previous coffee farms or secondary (i.e. disturbed) forest than in farms converted from primary (i.e. old growth) forest (A), but bacterial richness is not affected by previous land use (B). In organically managed farms, both fungal (C) and bacterial (D) richness were negatively correlated to the age of the coffee plants, and positively correlated to average tree cover (E and F, respectively). Circles in A-D are the average values for the farm. Kruskal–Wallis tests in A-B were used to determine global significance (bottom left of each panel) and Wilcoxon rank sum tests were used to test pairwise comparisons. In C-F, Pearson correlations were evaluated. The black line indicates a linear regression, and the grey-shaded area is the 95% confidence interval for the estimate. Figure 3. Open in new tabDownload slide Soil microbial richness varies according to environmental parameters. Fungal richness is significantly higher in farms converted from previous coffee farms or secondary (i.e. disturbed) forest than in farms converted from primary (i.e. old growth) forest (A), but bacterial richness is not affected by previous land use (B). In organically managed farms, both fungal (C) and bacterial (D) richness were negatively correlated to the age of the coffee plants, and positively correlated to average tree cover (E and F, respectively). Circles in A-D are the average values for the farm. Kruskal–Wallis tests in A-B were used to determine global significance (bottom left of each panel) and Wilcoxon rank sum tests were used to test pairwise comparisons. In C-F, Pearson correlations were evaluated. The black line indicates a linear regression, and the grey-shaded area is the 95% confidence interval for the estimate. Beta diversity Compositional differences in the fungal and bacterial communities were assessed across farms by looking at the Bray–Curtis distances between samples. Fungal and bacterial communities were strongly related to each other (Mantel's r = 0.466, P = 0.001). Adonis analysis showed that both fungal and bacterial communities clustered significantly according to whether they were organically or conventionally managed; the farm's previous land use, and the age of the coffee plants (P < 0.01 for all comparisons, Table 1). Among organic farms, bacterial community composition was related to the soil's chemical composition (Mantel's r = 0.33, P = 0.003). In contrast, fungal communities were significantly related to tree composition (Mantel's r = 0.17, P = 0.03). Table 1. Results of adonis analyses. Fungal and bacterial community matrixes were Hellinger-transformed to Bray–Curtis distances prior to analyses. The models did not include interactive terms, and significance was confirmed with 999 permutations. For all possible term orders, P < 0.03 for all terms, for both Fungi and Bacteria. . . Df . SumOfSqs . R2 . F . Pr(>F) . Fungi Previous land use 2 0.8652 0.05593 1.356 0.001 Management 1 0.4338 0.02804 1.3597 0.005 Coffee age 1 0.4512 0.02917 1.4142 0.001 Residual 43 13.7185 0.88686 Total 47 15.4686 1 Bacteria Previous land use 2 0.3455 0.05619 1.3711 0.004 Management 1 0.1861 0.03027 1.4775 0.006 Coffee age 1 0.1996 0.03246 1.5844 0.005 Residual 43 5.4175 0.88107 Total 47 6.1488 1 . . Df . SumOfSqs . R2 . F . Pr(>F) . Fungi Previous land use 2 0.8652 0.05593 1.356 0.001 Management 1 0.4338 0.02804 1.3597 0.005 Coffee age 1 0.4512 0.02917 1.4142 0.001 Residual 43 13.7185 0.88686 Total 47 15.4686 1 Bacteria Previous land use 2 0.3455 0.05619 1.3711 0.004 Management 1 0.1861 0.03027 1.4775 0.006 Coffee age 1 0.1996 0.03246 1.5844 0.005 Residual 43 5.4175 0.88107 Total 47 6.1488 1 Open in new tab Table 1. Results of adonis analyses. Fungal and bacterial community matrixes were Hellinger-transformed to Bray–Curtis distances prior to analyses. The models did not include interactive terms, and significance was confirmed with 999 permutations. For all possible term orders, P < 0.03 for all terms, for both Fungi and Bacteria. . . Df . SumOfSqs . R2 . F . Pr(>F) . Fungi Previous land use 2 0.8652 0.05593 1.356 0.001 Management 1 0.4338 0.02804 1.3597 0.005 Coffee age 1 0.4512 0.02917 1.4142 0.001 Residual 43 13.7185 0.88686 Total 47 15.4686 1 Bacteria Previous land use 2 0.3455 0.05619 1.3711 0.004 Management 1 0.1861 0.03027 1.4775 0.006 Coffee age 1 0.1996 0.03246 1.5844 0.005 Residual 43 5.4175 0.88107 Total 47 6.1488 1 . . Df . SumOfSqs . R2 . F . Pr(>F) . Fungi Previous land use 2 0.8652 0.05593 1.356 0.001 Management 1 0.4338 0.02804 1.3597 0.005 Coffee age 1 0.4512 0.02917 1.4142 0.001 Residual 43 13.7185 0.88686 Total 47 15.4686 1 Bacteria Previous land use 2 0.3455 0.05619 1.3711 0.004 Management 1 0.1861 0.03027 1.4775 0.006 Coffee age 1 0.1996 0.03246 1.5844 0.005 Residual 43 5.4175 0.88107 Total 47 6.1488 1 Open in new tab To disentangle the drivers of the microbial community, backwards and forwards selection were used to choose the most relevant tree canopy species, soil chemical components and management parameters for fungal and bacterial communities. For fungal communities, calcium, phosphorus, potassium, iron and soil carbon were identified as relevant chemical parameters; Rinorea squamata, Psidium spp., Manilkara zapota and an unidentified species locally known as ‘Cocotillo’ as relevant canopy tree species; and the farm's previous land use as a relevant management parameter. An RDA with these explanatory variables accounted for 32.4% of the variance in community composition. All explanatory variables were significant (P < 0.05 for all terms; P = 0.001 for the whole model, Fig. 4A). Figure 4. Open in new tabDownload slide A db-RDA of selected explanatory variables shows how different aspects of management affect soil fungal communities (A), and bacterial communities (B). Arrows for explanatory variables are colored according to the type of farm descriptor; the arrows corresponding to coffee as a previous land use (both panels) and conventional management (bottom panel) are omitted as they are linearly dependent on the other previous land use types. Circles and triangles represent plots. The db-RDA was significant (P < 0.001), and the significance of each term and axis is indicated with asterisks (*** P < 0.001; **P < 0.01; * P < 0.05). The tree species locally known as ‘Cocotillo’ could not be identified. Figure 4. Open in new tabDownload slide A db-RDA of selected explanatory variables shows how different aspects of management affect soil fungal communities (A), and bacterial communities (B). Arrows for explanatory variables are colored according to the type of farm descriptor; the arrows corresponding to coffee as a previous land use (both panels) and conventional management (bottom panel) are omitted as they are linearly dependent on the other previous land use types. Circles and triangles represent plots. The db-RDA was significant (P < 0.001), and the significance of each term and axis is indicated with asterisks (*** P < 0.001; **P < 0.01; * P < 0.05). The tree species locally known as ‘Cocotillo’ could not be identified. For the bacterial communities, backwards and forwards selection identified calcium, phosphorus, manganese, aluminium and potassium as relevant chemical parameters; Manilkara zapota, Ocotea spp., Inga oerstediana and ‘Cocotillo’ as relevant canopy tree species; and the farm's previous land use and whether it was organically or conventionally managed as relevant management parameters. An RDA with these explanatory variables accounted for 36.9% of the variance in the bacterial community composition. All explanatory variables were significant (P < 0.058 for all terms; P = 0.001 for the whole model, Fig. 4B). Fungal functional guilds Endophytes were the most abundant classified fungal functional group in conventional (22.9%) and organic (28.6%) coffee farms, but only comprised 13.3% of fungal communities in forest soils (Fig. 5A). The relative abundances of fungal plant pathogens varied with management as well, comprising 3.6% of the fungal community in conventional farms, and only 2.0% and 1.3% of organic and forest communities, respectively (Fig. 5A). When comparing functional guild composition across farms varying in time since established, endophytes and saprotrophs were the most abundant classified functional groups, with shifts in relative abundances with time since cultivation (Fig. 5B). To further illustrate the effects of management on fungal functional guilds, plant pathogen and endophyte community composition and their relative abundances were examined. Fungal plant pathogen community composition shifted across management, but with minor significance (PERMANOVA P = 0.085, R2 = 0.052, F model = 1.413); whereas, endophyte community composition did not shift significantly across management (PERMANOVA P = 0.271, R2 = 0.043, F model = 1.116) (Fig. 6A and B, respectively). Both fungal endophytes and plant pathogen composition shifted significantly across sites varying in previous land use (endophyte: PERMANOVA P = 0.022, R2 = 0.063, F model = 1.747; pathogen PERMANOVA P = 0.001, R2 = 0.082, F model = 2.337). Endophyte communities were dominated by Pleurotus and Rhizopycnis species across all samples regardless of management or previous land use. Figure 5. Open in new tabDownload slide Relative abundances of functional guilds of fungi across management (A), and years since cultivation (age of coffee plants, (B), as assigned by the FUNGuild program. Dashed line in panel A denotes relative abundance of 0.4, shifting the y-axis because >40% of fungal sequences across management categories were not classified by FUNGuild. Figure 5. Open in new tabDownload slide Relative abundances of functional guilds of fungi across management (A), and years since cultivation (age of coffee plants, (B), as assigned by the FUNGuild program. Dashed line in panel A denotes relative abundance of 0.4, shifting the y-axis because >40% of fungal sequences across management categories were not classified by FUNGuild. Figure 6. Open in new tabDownload slide Abundances of fungal genera assigned as plant pathogens (A), and endophytes (B), relative to the whole fungal communities detected in samples divided by management category. Figure 6. Open in new tabDownload slide Abundances of fungal genera assigned as plant pathogens (A), and endophytes (B), relative to the whole fungal communities detected in samples divided by management category. A PCA condensed soil parameters into axes representing gradients of macronutrient content, soil C and N and soil moisture; PC1 explained 31.11% of variation in soil chemistry across samples and runs from high mineral content (namely, Fe, Cu, B, P, Al and Zn; negative values) to higher moisture and %C and N (positive values, Fig. S3, Supporting Information). There was a significant effect of the first principal component on the relative abundances of plant pathogens (R2 = 0.19, P < 0.001), with a decline in pathogen abundance along the first PC axis (Fig. 7). There was no significant effect, however, of PC1 on the relative abundances of endophytes (R2 < 0.0001, P = 0.99). Figure 7. Open in new tabDownload slide Relationship between the relative abundances of plant pathogens and soil chemical parameters. Soil chemical parameters were summarized as the first axis of a principal component analysis constructed from soil chemical parameters (PC1). Points represent individual samples colored by management. PC1 runs from high mineral content (Fe, Cu, B, P, Al and Zn) to higher soil moisture, %C and %N. The principal component biplot is included in supplementary information. Figure 7. Open in new tabDownload slide Relationship between the relative abundances of plant pathogens and soil chemical parameters. Soil chemical parameters were summarized as the first axis of a principal component analysis constructed from soil chemical parameters (PC1). Points represent individual samples colored by management. PC1 runs from high mineral content (Fe, Cu, B, P, Al and Zn) to higher soil moisture, %C and %N. The principal component biplot is included in supplementary information. DISCUSSION The analysis of soils from 19 Nicaraguan coffee plantations subjected to a range of management regimes, showed that soil bacterial and fungal communities were altered by various aspects of management, which may have cascading effects on productivity. Previous research has also found significant effects of coffee management intensity on microbial composition (Hartmann et al. 2015). However, this study uncovered an added layer of complexity to microbial assembly processes across management intensity gradients, which was the previous land use of a plantation. The historical land use of a plantation affected the fungal and bacterial compartments in different ways, but demonstrated that agricultural intensification can interact with land use legacies to modulate soil microbial communities. The age of the coffee plantation was also a significant factor explaining variation in microbial composition, suggesting that assembly processes are dynamic and have temporal variability that cannot be inferred from sampling a farm at a single time point. Furthermore, it may take longer time periods of management changes to see parallel shifts in soil microbial communities. In conventionally-managed farms, soils received inputs of herbicides (i.e. Gramoxone) and industrial fertilizers (i.e. 15%N 15%P 15% K), while organically-managed plots received annual inputs of compost, bokashi or agricultural pulp. Correspondingly, a higher abundance of carbon and nitrogen was found in organically managed soils, and higher phosphorus in conventionally managed soils, in accordance with other studies of coffee agroecosystems in other countries (Partelli et al. 2012). Differences in soil carbon, phosphorus, and nitrogen are strong determinants of bacterial community composition (Delgado‐Baquerizo et al. 2017) and may explain how management affected the microbiomes. No relationship was detected between fungal or bacterial richness and canopy richness, consistent with previous research in grasslands (Prober et al. 2015), boreal forests (Tedersoo et al. 2016) which suggests that the relationship between below and above-ground richness may be context-dependent and modulated by other factors, such as evolutionary history (Goberna, Navarro-Cano and Verdú 2016) or local micro-environments (Tedersoo et al. 2016; Navrátilová et al. 2018). Indeed, in this study, other farm characteristics, including previous land use and the coffee stand age, were stronger determinants of the soil microbiome than canopy diversity. The previous land use of each farm had a strong effect on the soil microbiome, consistent with experimental findings of legacy effects on the assembly of soil microbial communities (Jurburg et al. 2017; Peerawat et al. 2018; Turley et al. 2020). This study further shows the persistent effect of previous land uses on soils under active cultivation. Fungal richness was strongly dependent on the land's previous land use, and was lowest in farms which had been converted from primary forest, suggesting that many forest-associated taxa do not persist in converted coffee plantations. In contrast, no relationship between bacterial richness and previous land use was detected. The higher sensitivity of fungal richness to agricultural land use change has been experimentally shown in restored agricultural fields compared to restored non-agricultural lands (Turley et al. 2020). Long-term (∼50 years) legacy effects of land use change on fungal communities have also been reported in tropical forests converted to oil palm plantations (McGuire et al. 2015) as well as in harvested temperate forests (Hartmann et al. 2015). In contrast, a meta-analysis of soil bacterial responses to land use change in the tropics found that bacterial alpha diversity increases following conversion to agricultural land (Petersen, Meyer and Bohannan 2019), possibly due to bacterial communities tracking changes in soil properties such as pH rather than responding directly to plant diversity loss following agricultural conversion. These results further support that fungal communities are more sensitive to land use change than bacterial communities, possibly due to the fact that fungi tend to form specialized associations with specific tree species, through differences in root exudates or leaf litter (De Deyn and Van Der Putten 2005; Prescott and Grayston 2013; Van der Putten et al. 2013; Barberán et al. 2015). Fungal communities respond more strongly than bacteria to changes in root exudates that result from altered plant diversity (Eisenhauer et al. 2017), which often results from land use change. Indeed, fungal community composition was strongly related to tree canopy composition in coffee agroecosystems. The legacy effects of previous land use held the strongest explanatory power for soil moisture and soil chemistry, reflecting how prior land use results in long-term alterations to the soil environment, which may directly or indirectly affect microbial community composition. The richness of bacteria, in contrast, was more strongly affected by the age of the coffee plants than other factors analysed in this study, decreasing linearly with coffee age. A similar, negative relationship between plant age and soil bacterial richness was reported for rubber trees (Peerawat et al. 2018), but to our knowledge, this experiment is the first to report such a relationship in the coffee agroecosystem. Slight shifts in management frequency (Peerawat et al. 2018) and changes in plant phenology with stand age may explain this pattern, but these two are interlinked and cannot be easily resolved, as management often depends on phenology. The production of coffee beans begins approximately three years after seed germination, and while the plant may stay productive for up to 80 years, the economic lifespan of the C. arabica plant is approximately 30 years (Wintgens 2009). In this study, the oldest plants were 45 years of age, after which farmers reported cutting down and reseeding the plots. No long-term changes in management with stand age were reported by the farmers in this study; however, it is likely that as plants matured, the frequency of management increased as harvesting became possible. Like bacterial richness, the concentration of calcium and potassium decreased with plant age. Although these were not correlated to each other, both calcium and potassium were strong drivers of bacterial community composition. Importantly, both fungal and bacterial richness decreased over time. Thus, it is possible that the gradual decrease in richness is a result of the continued pressures of management on the soil microbiota. The fungal functional guild analyses revealed that endophytic fungi made up significant proportions of the fungal communities across all coffee farms regardless of management, previous land use or time since cultivation. Fungal genera categorized as endophytes are generally nonpathogenic endosymbionts, forming symbioses with host plants in which endophytes can provide protective benefits against pathogenic fungi, or simply coexist as commensals (Wilson 1995). However, the nature of the relationship between plants and endophytes can vary under different environmental contexts, and many endophytes may be classified as latent pathogens or saprotrophs during some life stages (Carroll 1988). In some plants, endophytes may serve as beneficial endosymbionts through facilitating tolerance to herbivory, drought and heavy metal stress (Porras-Alfaro and Bayman 2011), which may be the case for Coffea plants. These findings are consistent with previous studies documenting abundant and diverse communities of endophytic fungi in coffee agroecosystems, but most studies examined fungal endophytes in the phyllosphere of Coffea plants (Vega et al. 2010; Oliveira et al. 2014; Saucedo-García et al. 2014). The relative abundances of endophytes across management categories in this study showed a trend of decreasing relative abundances of endophytes under conventional management compared to organic. Further, these results show a trend in the opposite direction for fungal plant pathogen communities, in which there was higher relative abundance of plant pathogens in conventionally managed soils compared to organic. This finding is likely due to a combination of biotic and abiotic drivers of fungal community assembly. First, interactions between populations of endophytes and pathogens may be driven by biotic interactions in which endophytes protect plants against antagonistic fungi known to cause disease; and these interactions have been briefly examined in coffee systems (Vega et al. 2010). Indeed, the most abundant endophytic fungi in this data set belong to the genus Mortierella (and other unidentified Mortierellaceae) which are symbiotrophic-saprotrophic fungi that are abundant in agricultural soils, particularly those under organic management (Zhang et al. 2020). Further, there is evidence that Mortierella species may act as beneficial root endosymbionts promoting plant growth, stabilizing soil nutrients and interacting synergistically with arbuscular mycorrhizal fungi in the rhizosphere (Li et al. 2018; Zhang et al. 2020). Another way in which management may drive shifts in the relative abundances of pathogenic fungal communities is through differences in abiotic conditions. The relationship found between relative abundances of fungi that cause disease in plants, and the soil chemical environment may be indirectly driven by management through fertilizer-induced shifts in the soil nutrient economy. This study highlights how multiple aspects of management alter the soil microbial communities in coffee agroecosystems. Previously, one study has shown how management affects soil nutrient availability and total microbial activity (Partelli et al. 2012), while another has found that microbial activity and biomass are better predictors of coffee productivity than soil chemistry (Silva Aragão et al. 2020). Similarly, the soil microbiota covaried with several factors, including management, but most notably, previous land use and stand age. While nutrient cycling rates were not measured, multiple studies have shown that biochemical processing rates, and decomposition in particular depend on the composition of the soil microbial community (Hartmann et al. 2015; Glassman et al. 2018; Ren et al. 2018). Thus, despite rarely being considered, it is likely that previous land use and stand age play an equally – if not more – important role in determining soil fertility for coffee production and the long-term sustainability of coffee agroecosystems. ACKNOWLEDGEMENTS We would like to thank A Gentile, C Dean and S Tem for their help and thoughtful discussions and M Mark and C Gillikin for help with permit acquisition. FUNDING This work was funded by the Jerry Bishop Environmental Scholarship, South Shore Audubon Society, NY; the Class of 1939 Summer Research Fellowship, Columbia University, NY and the German Centre for Integrative Biodiversity Research (iDiv) Halle-Jena-Leipzig, funded by the German Research Foundation (DFG FZT 118). Conflicts of interests None declared. REFERENCES Babbar LI , Zak DR. Nitrogen loss from coffee agroecosystems in Costa Rica: leaching and denitrification in the presence and absence of shade trees . J Environ Qual . 1995 ; 24 : 227 – 33 . Google Scholar Crossref Search ADS WorldCat Bacon CM , Gliessman SR, Mendez VE et al. Confronting the Coffee Crisis: Fair Trade, Sustainable Livelihoods and Ecosystems in Mexico and Central America , Cambridge, MA, U.S.A. : MIT Press , 2008 . Google Scholar Crossref Search ADS Google Scholar Google Preview WorldCat COPAC Bacon CM . A spot of coffee in crisis: Nicaraguan smallholder cooperatives, fair trade networks, and gendered empowerment . Lat Am Perspect . 2010 ; 37 : 50 – 71 . Google Scholar Crossref Search ADS WorldCat Barberán A , Mcguire KL, Wolf JA et al. Relating belowground microbial composition to the taxonomic, phylogenetic, and functional trait distributions of trees in a tropical forest . Ecol Lett . 2015 ; 18 : 1397 – 405 . Google Scholar Crossref Search ADS PubMed WorldCat Beer J , Muschler R, Kass D et al. Shade management in coffee and cacao plantations . Agrofor Syst . 1997 ; 38 : 139 – 64 . Google Scholar Crossref Search ADS WorldCat Bellemain E , Carlsen T, Brochmann C et al. ITS as an environmental DNA barcode for fungi: an in silico approach reveals potential PCR biases . BMC Microbiol . 2010 ; 10 : 189 . Google Scholar Crossref Search ADS PubMed WorldCat Berg G . Plant–microbe interactions promoting plant growth and health: perspectives for controlled use of microorganisms in agriculture . Appl Microbiol Biotechnol . 2009 ; 84 : 11 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat Brussaard L , De Ruiter PC, Brown GG. Soil biodiversity for agricultural sustainability . Agric Ecosyst Environ . 2007 ; 121 : 233 – 44 . Google Scholar Crossref Search ADS WorldCat Buechley ER , Şekercioğlu ÇH, Atickem A et al. Importance of Ethiopian shade coffee farms for forest bird conservation . Biol Conserv . 2015 ; 188 : 50 – 60 . Google Scholar Crossref Search ADS WorldCat Caporaso JG , Kuczynski J, Stombaugh J et al. QIIME allows analysis of high-throughput community sequencing data . Nat Methods . 2010 ; 7 : 335 – 6 . Google Scholar Crossref Search ADS PubMed WorldCat Caporaso JG , Lauber CL, Walters WA et al. Ultra-high-throughput microbial community analysis on the Illumina HiSeq and MiSeq platforms . ISME J . 2012 ; 6 : 1621 – 4 . Google Scholar Crossref Search ADS PubMed WorldCat Carroll G . Fungal endophytes in stems and leaves: from latent pathogen to mutualistic symbiont . Ecology . 1988 ; 69 : 2 – 9 . Google Scholar Crossref Search ADS WorldCat Chaparro JM , Sheflin AM, Manter DK et al. Manipulating the soil microbiome to increase soil health and plant fertility . Biol Fertil Soils . 2012 ; 48 : 489 – 99 . Google Scholar Crossref Search ADS WorldCat De Deyn GB , Van Der Putten WH. Linking aboveground and belowground diversity . Trends Ecol Evol . 2005 ; 20 : 625 – 33 . Google Scholar Crossref Search ADS PubMed WorldCat Delgado‐Baquerizo M , Reich PB, Khachane AN et al. It is elemental: soil nutrient stoichiometry drives bacterial diversity . Environ Microbiol . 2017 ; 19 : 1176 – 88 . Google Scholar Crossref Search ADS PubMed WorldCat DeSantis TZ , Hugenholtz P, Larsen N et al. Greengenes, a chimera-checked 16S rRNA gene database and workbench compatible with ARB . Appl Environ Microbiol . 2006 ; 72 : 5069 – 72 . Google Scholar Crossref Search ADS PubMed WorldCat Donald PF . Biodiversity impacts of some agricultural commodity production systems . Conserv Biol . 2004 ; 18 : 17 – 38 . Google Scholar Crossref Search ADS WorldCat Edgar RC . UPARSE: highly accurate OTU sequences from microbial amplicon reads . Nat Meth . 2013 ; 10 : 996 – 8 . Google Scholar Crossref Search ADS WorldCat Eisenhauer N , Lanoue A, Strecker T et al. Root biomass and exudates link plant diversity with soil bacterial and fungal biomass . Sci Rep . 2017 ; 7 : 44641 . Google Scholar Crossref Search ADS PubMed WorldCat Falkowski PG , Fenchel T, Delong EF. The microbial engines that drive Earth's biogeochemical cycles . Science (80-) . 2008 ; 320 : 1034 – 9 . Google Scholar Crossref Search ADS WorldCat FAO . FAOSTAT statistics database . 2017 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Fichtner A , Von Oheimb G, Härdtle W et al. Effects of anthropogenic disturbances on soil microbial communities in oak forests persist for more than 100 years . Soil Biol Biochem . 2014 ; 70 : 79 – 87 . Google Scholar Crossref Search ADS WorldCat Gardes M , Bruns TD. ITS primers with enhanced specificity for basidiomycetes‐application to the identification of mycorrhizae and rusts . Mol Ecol . 1993 ; 2 : 113 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat Glassman SI , Weihe C, Li J et al. Decomposition responses to climate depend on microbial community composition . Proc Natl Acad Sci USA . 2018 ; 115 : 11994 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat Goberna M , Navarro-Cano JA, Verdú M. Opposing phylogenetic diversity gradients of plant and soil bacterial communities . Proc R Soc B Biol Sci . 2016 ; 283 : 20153003 . Google Scholar Crossref Search ADS WorldCat Goodall KE , Bacon CM, Mendez VE. Agriculture, Ecosystems and Environment Shade tree diversity, carbon sequestration, and epiphyte presence in coffee agroecosystems : A decade of smallholder management in San . Agriculture, Ecosyst Environ . 2014 ; 199 : 200 – 6 . Google Scholar Crossref Search ADS WorldCat Hallama M , Pekrun C, Lambers H et al. Hidden miners–the roles of cover crops and soil microorganisms in phosphorus cycling through agroecosystems . Plant Soil . 2019 ; 434 : 7 – 45 . Google Scholar Crossref Search ADS WorldCat Hartmann M , Frey B, Mayer J et al. Distinct soil microbial diversity under long-term organic and conventional farming . ISME J . 2015 ; 9 : 1177 – 94 . Google Scholar Crossref Search ADS PubMed WorldCat Hernández-Martínez G , Manson RH, Hernández AC. Quantitative classification of coffee agroecosystems spanning a range of production intensities in central Veracruz, Mexico . Agric Ecosyst Environ . 2009 ; 134 : 89 – 98 . Google Scholar Crossref Search ADS WorldCat Hobbs RJ , Arico S, Aronson J et al. Novel ecosystems: Theoretical and management aspects of the new ecological world order . Glob Ecol Biogeogr . 2006 ; 15 : 1 – 7 . Google Scholar Crossref Search ADS WorldCat Jurburg SD , Nunes I, Brejnrod A et al. Legacy effects on the recovery of soil bacterial communities from extreme temperature perturbation . Front Microbiol . 2017 ; 8 : 1832 . Google Scholar Crossref Search ADS PubMed WorldCat Komar O . Priority contribution. Ecology and conservation of birds in coffee plantations: A critical review . Bird Conserv Int . 2006 ; 16 : 1 – 23 . Google Scholar Crossref Search ADS WorldCat Kulmatiski A , Beard KH. Decoupling plant-growth from land-use legacies in soil microbial communities . Soil Biol Biochem . 2008 ; 40 : 1059 – 68 . Google Scholar Crossref Search ADS WorldCat Kumsa L , Hylander K, Gurmessa D et al. Patch area and current coffee management determine woody plant diversity in patches of semi-forest coffee embedded in an agricultural matrix . Glob Ecol Conserv . 2016 ; 8 : 230 – 40 . Google Scholar Crossref Search ADS WorldCat Lauber CL , Hamady M, Knight R et al. Pyrosequencing-based assessment of soil pH as a predictor of soil bacterial community structure at the continental scale . Appl Environ Microbiol . 2009 ; 75 : 5111 – 20 . Google Scholar Crossref Search ADS PubMed WorldCat Li F , Chen L, Redmile‐Gordon M et al. Mortierella elongata's roles in organic agriculture and crop growth promotion in a mineral soil . L Degrad Dev . 2018 ; 29 : 1642 – 51 . Google Scholar Crossref Search ADS WorldCat Ma H , Pineda A, Hannula SE et al. Steering root microbiomes of a commercial horticultural crop with plant-soil feedbacks . Appl Soil Ecol . 2020 ; 150 : 103468 . Google Scholar Crossref Search ADS WorldCat McGuire KL , D'Angelo H, Brearley FQ et al. Responses of soil fungi to logging and oil palm agriculture in southeast asian tropical forests . Microb Ecol . 2015 ; 69 : 733 – 47 . Google Scholar Crossref Search ADS PubMed WorldCat McGuire KL , Payne SG, Palmer MI et al. Digging the New York City skyline: soil fungal communities in green roofs and city parks . PLoS One . 2013 ; 8 . OpenURL Placeholder Text WorldCat McMurdie PJ , Holmes S. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data . PLoS One . 2013 ; 8 : e61217 . Google Scholar Crossref Search ADS PubMed WorldCat Munroe JW , Soto G, de E et al. Soil microbial and nutrient properties in the rhizosphere of coffee under agroforestry management . Appl Soil Ecol . 2015 ; 93 : 40 – 6 . Google Scholar Crossref Search ADS WorldCat Navrátilová D , Tláskalová P, Kohout P et al. Diversity of fungi and bacteria in species-rich grasslands increases with plant diversity in shoots but not in roots and soil . FEMS Microbiol Ecol . 2018 ; 95 , DOI: 10.1093/femsec/fiy208 . OpenURL Placeholder Text WorldCat Nemergut DR , Schmidt SK, Fukami T et al. Patterns and Processes of Microbial Community Assembly . Microbiol Mol Biol Rev . 2013 ; 77 : 342 – 56 . Google Scholar Crossref Search ADS PubMed WorldCat Nguyen NH , Song Z, Bates ST et al. FUNGuild: an open annotation tool for parsing fungal community datasets by ecological guild . Fungal Ecol . 2016 ; 20 : 241 – 8 . Google Scholar Crossref Search ADS WorldCat Oksanen J , Blanchet FG, Kindt R et al. The vegan package . Community Ecol . 2007 . OpenURL Placeholder Text WorldCat Oliveira R , Souza R, Lima T et al. Endophytic fungal diversity in coffee leaves (Coffea arabica) cultivated using organic and conventional crop management systems . Mycosphere . 2014 ; 5 : 523 – 30 . Google Scholar Crossref Search ADS WorldCat Partelli FL , Vieira HD, de Brito Ferreira EP et al. Chemical and microbiological soil characteristics under conventional and organic coffee production systems . Commun Soil Sci Plant Anal . 2012 ; 43 : 847 – 64 . Google Scholar Crossref Search ADS WorldCat Peay KG , Baraloto C, Fine PVA. Strong coupling of plant and fungal community structure across western Amazonian rainforests . ISME J . 2013 ; 7 : 1852 – 61 . Google Scholar Crossref Search ADS PubMed WorldCat Peerawat M , Blaud A, Trap J et al. Rubber plantation ageing controls soil biodiversity after land conversion from cassava . Agric Ecosyst Environ . 2018 ; 257 : 92 – 102 . Google Scholar Crossref Search ADS WorldCat Perfecto I , Vandermeer J, Philpott SM. Complex ecological interactions in the coffee agroecosystem . Annu Rev Ecol Evol Syst . 2014 ; 45 : 137 – 58 . Google Scholar Crossref Search ADS WorldCat Petersen IAB , Meyer KM, Bohannan BJM. Meta-analysis reveals consistent bacterial responses to land use change across the tropics . Front Ecol Evol . 2019 ; 7 : 391 . Google Scholar Crossref Search ADS WorldCat Philpott SM , Arendt WJ, Armbrecht I et al. Biodiversity loss in Latin American coffee landscapes: review of the evidence on ants, birds, and trees . Conserv Biol . 2008 ; 22 : 1093 – 105 . Google Scholar Crossref Search ADS PubMed WorldCat Philpott SM . Changes in arboreal ant populations following pruning of coffee shade-trees in Chiapas, Mexico . Agrofor Syst . 2005 : 219 – 24 . OpenURL Placeholder Text WorldCat Porras-Alfaro A , Bayman P. Hidden fungi, emergent properties: endophytes and microbiomes . Annu Rev Phytopathol . 2011 ; 49 : 291 – 315 . Google Scholar Crossref Search ADS PubMed WorldCat Prescott CE , Grayston SJ. Tree species influence on microbial communities in litter and soil: current knowledge and research needs . For Ecol Manage . 2013 ; 309 : 19 – 27 . Google Scholar Crossref Search ADS WorldCat Prober SM , Leff JW, Bates ST et al. Plant diversity predicts beta but not alpha diversity of soil microbes across grasslands worldwide . Ecol Lett . 2015 ; 18 : 85 – 95 . Google Scholar Crossref Search ADS PubMed WorldCat Ren C , Wang T, Xu Y et al. Differential soil microbial community responses to the linkage of soil organic carbon fractions with respiration across land-use changes . For Ecol Manage . 2018 ; 409 : 170 – 8 . Google Scholar Crossref Search ADS WorldCat Ricketts TH , Daily GC, Ehrlich PR et al. Economic value of tropical forest to coffee production . Proc Natl Acad Sci . 2004 ; 101 : 12579 – 82 . Google Scholar Crossref Search ADS PubMed WorldCat Saucedo-García A , Anaya AL, Espinosa-García FJ et al. Diversity and communities of foliar endophytic fungi from different agroecosystems of Coffea arabica L. in two regions of Veracruz, Mexico . PLoS One . 2014 ; 9 . OpenURL Placeholder Text WorldCat Schrama M , Vandecasteele B, Carvalho S et al. Effects of first‐and second‐generation bioenergy crops on soil processes and legacy effects on a subsequent crop . gCb bioenergy . 2016 ; 8 : 136 – 47 . Google Scholar Crossref Search ADS WorldCat Silva Aragão OO , de Oliveira-Longatti SM, Santos de Castro Caputo P et al. Microbiological indicators of soil quality are related to greater coffee yield in the Brazilian Cerrado region . Ecol Indic . 2020 ; 113 : 106205 . Google Scholar Crossref Search ADS WorldCat Six J , Frey SD, Thiet RK et al. Bacterial and fungal contributions to carbon sequestration in agroecosystems . Soil Sci Soc Am J . 2006 ; 70 : 555 – 69 . Google Scholar Crossref Search ADS WorldCat Tedersoo L , Bahram M, Cajthaml T et al. Tree diversity and species identity effects on soil fungi, protists and animals are context dependent . ISME J . 2016 ; 10 : 346 – 62 . Google Scholar Crossref Search ADS PubMed WorldCat Thompson LR , Sanders JG, McDonald D et al. A communal catalogue reveals Earth's multiscale microbial diversity . Nature . 2017 . OpenURL Placeholder Text WorldCat Toju H , Peay KG, Yamamichi M et al. Core microbiomes for sustainable agroecosystems . Nat Plants . 2018 ; 4 : 247 – 57 . Google Scholar Crossref Search ADS PubMed WorldCat Tscharntke T , Clough Y, Bhagwat SA et al. Multifunctional shade‐tree management in tropical agroforestry landscapes–a review . J Appl Ecol . 2011 ; 48 : 619 – 29 . Google Scholar Crossref Search ADS WorldCat Turley NE , Bell-Dereske L, Evans SE et al. Agricultural land-use history and restoration impact soil microbial biodiversity . J Appl Ecol . 2020 : 852 – 63 . OpenURL Placeholder Text WorldCat Vaast P , Bertrand B, Perriot J et al. Fruit thinning and shade improve bean characteristics and beverage quality of coffee (Coffea arabica L.) under optimal conditions . 2006 ; 204 : 197 – 204 . Van der Putten WH , Bardgett RD, Bever JD et al. Plant–soil feedbacks: the past, the present and future challenges . J Ecol . 2013 ; 101 : 265 – 76 . Google Scholar Crossref Search ADS WorldCat van der Putten WH , Bradford MA, Pernilla Brinkman E et al. Where, when and how plant–soil feedback matters in a changing world . Funct Ecol . 2016 ; 30 : 1109 – 21 . Google Scholar Crossref Search ADS WorldCat Vega FE , Simpkins A, Aime MC et al. Fungal endophyte diversity in coffee plants from Colombia, Hawai'i, Mexico and Puerto Rico . fungal Ecol . 2010 ; 3 : 122 – 38 . Google Scholar Crossref Search ADS WorldCat Vukicevich E , Lowery DT, Úrbez-Torres JR et al. Groundcover management changes grapevine root fungal communities and plant-soil feedback . Plant Soil . 2018 ; 424 : 419 – 33 . Google Scholar Crossref Search ADS WorldCat Wang Q , Garrity GM, Tiedje JM et al. Naive Bayesian classifier for rapid assignment of rRNA sequences into the new bacterial taxonomy . Appl Environ Microbiol . 2007 ; 73 : 5261 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat Wardle DA , Bardgett RD, Klironomos JN et al. Ecological linkages between aboveground and belowground biota . Science (80-) . 2004 ; 304 : 1629 – 33 . Google Scholar Crossref Search ADS WorldCat Wickham H . ggplot2 . Wiley Interdiscip Rev Comput Stat . 2011 ; 3 : 180 – 5 . Google Scholar Crossref Search ADS WorldCat Wilson D . Endophyte: the evolution of a term, and clarification of its use and definition . Oikos . 1995 : 274 – 6 . OpenURL Placeholder Text WorldCat Wintgens JN . The coffee plant . Coffee growing, Process Sustain Prod A Guideb Grow Process traders Res . 2009 : 3 – 24 . OpenURL Placeholder Text WorldCat Zhang K , Bonito G, Hsu C-M et al. Mortierella elongata Increases Plant Biomass among Non-Leguminous Crop Species . Agronomy . 2020 ; 10 : 754 . Google Scholar Crossref Search ADS WorldCat © FEMS 2020. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - Soil microbial composition varies in response to coffee agroecosystem management
JF - FEMS Microbiology Ecology
DO - 10.1093/femsec/fiaa164
DA - 2020-09-01
UR - https://www.deepdyve.com/lp/oxford-university-press/soil-microbial-composition-varies-in-response-to-coffee-agroecosystem-od3KHuxGCb
VL - 96
IS - 9
DP - DeepDyve
ER -