Ecological spillover of coffee-associated foliar fungi from field to forest as a biotic edge effect in the Neotropics

Ecological spillover of coffee-associated foliar fungi from field to forest as a biotic edge effect in the Neotropics | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Biological Sciences - Article Ecological spillover of coffee-associated foliar fungi from field to forest as a biotic edge effect in the Neotropics Jeffrey Lackmann, Benedicte Bachelot, Catherine Lindell, Elena Prado-Ragan, and 7 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7781221/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted You are reading this latest preprint version Abstract Agricultural expansion creates mosaic landscapes where managed land borders undisturbed ecosystems, facilitating the ecological spillover of organisms between habitats. While fungal spillover from natural systems to crops is relatively well-studied, the reverse—agricultural fungi colonizing adjacent forests—remains underexplored. We investigated whether coffee fields act as reservoirs for foliar fungi that spill over into adjacent tropical forests, particularly affecting close relatives of coffee (Rubiaceae). Using transect-based field surveys, airborne spore sampling, and metabarcoding of fungal communities, we found that leaf spot incidence was much higher in coffee than in adjacent understory forest leaves and declined with distance into the forest for Rubiaceae plants. Pleosporales was the dominant fungal order in coffee leaves and declined in abundance with distance from the forest edge in both airborne samples and Rubiaceae leaves, suggesting propagule pressure from coffee fields led to changes in the phyllobiome of close coffee relatives in the adjacent forest. Among the 1,000 most abundant coffee-associated fungal taxa, 19 declined with distance into the forest, and 93 were more abundant on phylogenetically close relatives of coffee compared to non-Rubiaceae plants. Several of the closest matches for these taxa were potential plant pathogens. These results demonstrate that high-density crops can act as reservoirs for fungi capable of colonizing adjacent forest plants, disproportionately affecting plants closely related to the crop. This appears to disrupt natural plant pathogen dynamics in forests such that close relatives of the crop plant near the crop-forest border face increased disease pressure and may be at a competitive disadvantage as a result. These findings highlight an underappreciated edge effect of agriculture in fragmented landscapes. Earth and environmental sciences/Ecology/Forest ecology Earth and environmental sciences/Ecology/Agroecology Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Introduction Foliar fungi inhabit leaf surfaces and interiors, exhibit high dispersal potential (Golan & Pringle, 2017 ), varying degrees of host specificity (Apigo & Oono, 2022 ), and distinct functional roles as both pathogens and mutualists capable of mediating plant community structure (Pajares-Murgó et al., 2024 ). The widespread conversion of natural ecosystems to agriculture creates mosaic landscapes where managed land directly borders wildlands (Pendrill et al., 2022 ; Ma et al., 2023 ). Edges between managed and natural communities create ecological interfaces where foliar fungi from both communities encounter each other. These fungi and their associated functions in the ecosystem can move across these boundaries in a phenomenon known as ecological spillover (Blitzer et al., 2012 ; Spear et al., 2018 ). Ecotones between agricultural and natural systems are often characterized by pronounced edge effects, where environmental conditions, species composition, and biotic interactions differ markedly from the ecosystem interiors (Rice, 1999 ; Cadenasso et al., 2003 ). In tropical forest landscapes fragmented by agriculture, edges are known to alter microclimate, species turnover, and ecological processes (Benítez-Malvido et al., 2018 ), including the spread of fungal pathogens (Johnson & Haddad, 2011 ). High-density monocultures like coffee plantations generate steep biotic gradients at their borders, where elevated abundances of host plants and associated pathogens may increase disease pressure in adjacent natural communities (Laurance et al., 2007 ). While many studies have focused on natural systems as reservoirs of organisms that affect nearby agricultural systems, much less is known about the extent to which organisms from agriculture disperse into and modify adjacent natural systems (Blitzer et al., 2012 ; Reis Medeiros et al., 2022 ). Many foliar fungi associate closely with managed systems and may accumulate at high enough densities to spill into adjacent wildlands (Gilbert & Hubbell, 1996 ). In forests, Janzen–Connell effects—driven by increasing enemy pressure at higher conspecific densities—limit dominant species and promote the persistence of rare species (Janzen, 1970; Connell, 1971; Liu et al., 2012; Bayandala et al., 2017). These same density-dependent processes can also drive disease outbreaks in managed crop systems when conditions favor host-associated pathogens (Hopkins et al., 2020; Pfäffle et al., 2015; Wang et al., 2021). Crops grown at high density can amplify populations of pathogens, pests, and natural enemies, increasing the potential for ecological spillover into adjacent natural systems. The likelihood of spillover also depends on host relatedness: foliar fungi are more likely to establish in nearby habitats when they encounter compatible hosts (Gilbert & Webb, 2007 ). While some fungi (e.g., Xylaria spp.) have broad host ranges, others (e.g., Colletotrichum spp.) specialize on particular plant lineages (Arnold & Lutzoni, 2007; Qian et al., 2018; Yang et al., 2023; Guo et al., 2024). Thus, plants in natural systems closely related to a crop species should be more susceptible to colonization by crop-associated fungi than more distantly related neighbors (Parker et al., 2015; Kembel & Mueller, 2014; Chen et al., 2022). If these fungi influence host fitness, they could alter plant community composition over time, favoring close relatives of the crop if they act as mutualists or selecting against them if pathogenic (Arnold et al., 2003). The foliar fungal communities of coffee plantations in Central America provide an ecologically and economically relevant tropical system to assess whether ecological spillover of crop-associated fungi occurs and whether nearby close relatives of the host plant are more likely to be colonized than more distantly related plant species. Coffee ( Coffea arabica L., Family Rubiaceae), as one of the most economically important crops in the tropics, plays a central role in shaping both rural livelihoods and land use (Harvey et al., 2021 ). These plantations are often managed as high-density monocultures in close proximity to remnant forest (Caudill et al., 2017 ). The Rubiaceae family includes many native forest understory species in Central America (Delprete & Jardim, 2012 ; Razafimandimbison & Rydin, 2024 ). This physical proximity to forests and phylogenetic overlap with many forest plants creates conditions where ecological spillover is likely to occur. We asked whether coffee plantations act as reservoirs for foliar fungi that can affect forest understory plants. We hypothesized that if coffee serves as a reservoir, we would detect coffee-associated fungi at higher abundances in the air and in the leaves of forest plants near the coffee-forest border compared to plants located further into the forest. Second, we hypothesized that if host relatedness affects colonization, coffee-associated fungi should occur with greater frequency on forest plants the more closely related they are to coffee. Methods Study sites Fieldwork was conducted March to August 2021 at eight sites in Coto Brus, Puntarenas Province, Costa Rica, near the Las Cruces Biological Station (8°47′7″N, 82°57′32″W), in a tropical premontane ecosystem at an elevational range of 800–1133 m asl with mean annual rainfall of 4000 mm and mean annual temperature of 21°C (Holdridge & Grenke, 1971; Zahawi et al., 2015 ). All sites contained privately owned coffee fields ranging in age from 10 to 67 yrs directly adjacent to primary forest > 10 ha. We established a 200 m transect extending 100 m into both the coffee field and the adjacent forest at seven sites ensuring that the 100 m mark was at least 100 m from any other edge of the habitat (Fig. 1 A). At the eighth site (RN), the transect extended only 50 m into the coffee to avoid sampling within 50 m of the opposite edge of the field. Leaf collection for foliar fungi To assess the spatial patterns of foliar fungi in coffee fields and the adjacent forest understory, using flame-sterilized scissors we removed a leaf from the second fully expanded leaf pair (counting from the branch tip) from four randomly selected coffee plants within a 5 m radius at 10, 25, 50, and 100 m away from the coffee-forest border at seven sites, while doubling the samples at 50 m in the eighth site, to keep the number of coffee leaf samples consistent across the sites (Fig. 1 ). We repeated this sampling process for understory forest plants at 10, 25, 50, and 100 m into the forest, sampling in a 5 m radius from the distance point along transect. To investigate whether forest plants shared more foliar fungi with coffee when closely than more distantly related, we sampled leaves from 2–6 individuals of each of five plant species (three Rubiaceae, two non-Rubiaceae) at each distance (Fig. 1 ; Table 1 ). If at least three individuals of a given species could not be located at a given distance, additional species within the relevant sample category (Rubiaceae or non-Rubiaceae) were sampled. We sampled the same forest plant species across distances and sites whenever possible but, because plant species diversity was so high, we often had to sample different plant species at different distances and sites (Table S1 ). In total, we sampled 131 coffee leaves (one leaf sample was lost prior to image analysis) and 807 forest leaves (Table 1 ) spanning 26 species and eight families (Table S1 ) from June 1-July 7th, 2021. Representative material from each forest plant species was collected, identified, and deposited in the herbaria at Las Cruces Biological Station and the National Herbarium of Costa Rica. Leaf imaging for damage morphotyping & seedling damage survey The same day that leaves were collected, they were transported to Las Cruces Biological Station in sterilized coin envelopes at ambient air temperature in sealed plastic bags partially filled with desiccant. At the field station, we photographed each leaf against a uniform background. We manually screened each image for the incidence of disease symptoms including leaf spot, powdery mildew, downy mildew, amorphous chlorosis, and amorphous necrosis. After leaves were photographed, sections of each leaf were haphazardly selected and 800 mg (fresh weight) was removed with a flame-sterilized scalpel, avoiding the midvein. Samples were dried in open 2-mL microcentrifuge tubes placed above silica gel in air-tight containers for ~ 12 hours and transported to the Aldrich-Wolfe lab (Fig. 1 ). Samples were lyophilized for 48 h at -80°C using a FreeZone 6 Plus freeze dryer (Labconco Corporation, Kansas City, MO, USA) and subsequently stored at ambient temperature. In late May 2022, we established ten 1 × 1 m plots over the existing seedling community in a subset of four of the eight forest sites, starting at 5 m from the forest edge and proceeding every 10 m to 105 m into the forest. We individually tagged all woody plant species in the plot, identified them to species, and scored them for leaf fungal damage (%). We assessed the amount of leaf fungal damage as a visual proportion of total leaf damage to undamaged leaf for the entire seedling. Foliar damage was re-measured bi-weekly and seedling mortality status was assessed weekly for eight weeks. If more than 20 individuals of any species were present in a given plot, all were counted but only a random 20 were tagged. Using these data, we calculated total amount of foliar damage at the plot level averaging across seedlings. We also calculated average phylogenetic distance from coffee at the plot level using the R package V.PhyloMaker2. Sampling airborne fungi To investigate airborne propagule pressure, Modified Wilson and Cook Samplers (hereafter “air samplers”; Goossens & Offer, 2000 ) were constructed at Las Cruces Biological Station. These wind-driven devices consist of a free-moving mast-mounted pole and aluminum sail that rotates the device with the wind. Each 1.36-m high air sampler was equipped with three wide-mouth 32-oz (946-mL) plastic jars (ULINE, #S-18072, Pleasant Prairie, Wisconsin, USA) positioned at 30, 80, and 130 cm above the soil surface to ensure sampling of air at heights in the typical range of understory shrub species. We fitted each jar with modified lids containing intake and outtake tubes for airflow oriented in relation to the sail such that the intake tube faced into the wind to capture airborne particles (Fig. S1 ). Prior to field use, collection jars and lids were sterilized in 10% bleach ( v/v ) for 30 min and dried for ~ 45 min at 60°C to evaporate any remaining bleach solution. After drying, the modified lids were sealed onto the jars with plastic wrap covering the intake and outtake tubes until placed in the field. To sample airborne fungi, we placed one air sampler each at 10, 25, 50, and 100 m from the coffee-forest border in the forest understory and at 50 m from the coffee-forest border in the coffee field (Fig. 1 ). To compare which propagules were moving from coffee to forest and vice versa, we also placed two air samplers at the coffee-forest border on fixed poles that could not rotate, one facing the coffee and the other facing the forest (Fig. 1 ). Each month, deployed jars were replaced with bleach-sterilized jars in the field, and the deployed jars were sealed, transported to Las Cruces Biological Station, refrigerated, and processed within 24 h. We rinsed the contents of the three collection jars from each air sampler with 50 mL of ultrapure water onto a 0.45 µm GN-6 nitrocellulose membrane (Metricel, St. Louis, MO, USA) on a Büchner funnel attached to a Welch WOB-L Fluid Aspiration Vacuum Pump (Chemtech Scientific, Tampa, FL, USA) to collect particulates including fungal propagules, which are generally larger than 1 µm (Golan & Pringle, 2017 ). This resulted in seven samples for each site each month, one for each air sampler, containing the airborne fungi accumulated over the course of the preceding month. Samples were collected at six of the sites in April 2021, at seven of the sites in May 2021, and at all eight sites in June and July 2021, resulting in a total of 203 airborne samples. We included negative controls by processing as above an unused collection jar sterilized and dried identically to those deployed in the field each collection day to determine if any fungi persisted during sterilization or were introduced during filtration. We placed each airborne sample in an individual, sterile 5-mL tube from the DNeasy® PowerWater Kit (Qiagen, Germantown MD, USA). We dried these samples by puncturing tube lids with a flame-sterilized needle, placing them inside a sealed plastic bag filled with silica gel to dry for 24 h at 4°C. Tubes were then sealed with new lids and stored at -20°C. Tubes were shipped frozen on dry ice to North Dakota State University and stored at -20°C prior to DNA isolation. Molecular characterization of leaf and airborne fungi For both leaf and air samples, we added a sterile 6.35 mm chrome-steel bead (BioSpec Products, Inc., Bartlesville, Oklahoma, USA) to each microcentrifuge tube. The desiccated leaf material was pulverized dry while the air samples were pulverized wet, per kit instructions, using a TissueLyser II (Qiagen, Germantown, Maryland, USA) at 27 Hz for two minutes for each plate orientation (4 min total) per operating instructions. DNA from the airborne samples was isolated using the DNeasy® PowerWater kit following the manufacturer’s protocol (Qiagen, Germantown MD, USA). We processed one DNA blank for every 47 samples during isolation. DNA from our leaf material was isolated using the Qiagen DNeasy® Plant Kit following the manufacturer’s protocol, with the following modifications. Because we had twice the maximum amount of dried sample in individual microcentrifuge tubes (~ 200 mg) for the 96-well format, we doubled the amounts of AP1 and RNase and added them directly to each sample in its microcentrifuge tube in the heating step to ensure samples were at the correct concentration for adequate cell lysis and to improve RNA degradation. After vortexing, we transferred 400 µL of each lysate to its position in the 96-well plates, thereby adjusting the sample to its optimal concentration and volume and allowing us to proceed with the manufacturer’s protocol. We included the optional third ethanol wash step to further remove impurities from DNA. The DNA extracts were eluted from the column twice with 50 µL of Tris-EDTA and stored at -80°C. For leaves, we pooled 20 µL of DNA extract from each individual of a given species at each distance within each site into a single composite sample for that species at that distance at that site. Unfortunately, during this process six samples were inadvertently destroyed, one coffee sample and five non-Rubiaceae samples (Table 1 ). Our negative control eluates were pooled as one sample prior to shipping. We shipped all eluates on dry ice to the University of Minnesota Genomic Center (UMGC, Minneapolis, MN, USA) for library preparation following the protocol in Sternhagen et al. ( 2020 ) and sequencing of the ITS2 region of the small ribosomal subunit using the fungal-specific primers ITS4 (White et al., 1990 ) and 5.8SR (Vilgalys & Hester, 1990 ) and Illumina MiSeq™ protocol in Gohl et al. ( 2016 ). UMGC conducted preliminary quality control and demultiplexing. We processed the raw FASTA sequences for the ITS2 region via the PIPITS 3.0 bioinformatics pipeline (Gweon et al., 2015 ) with a 97% similarity threshold, using the UNITE v8.3 database (Abarenkov et al., 2024 ). This fully automated pipeline prepares FASTA sequences, detects and removes chimeras and sequences < 100 bp in length, assigns operational taxonomic units (OTUs), designates a taxonomic prediction and a confidence score for that prediction based on known sequences from the UNITE fungal database, and generates an OTU abundance matrix. We chose not to rarefy the OTU table to avoid discarding data (McMurdie & Holmes, 2014 ). Only OTUs with a kingdom confidence score of 1 (the highest possible score) were included in the final dataset. OTUs with taxonomic designations that had confidence scores < 0.87 were considered unidentified. The highest read count for each OTU detected in the negative controls was subtracted from its read count, if any, in each of the biological samples. Table 1 Sampling effort across sites for image assessment of pathogen damage and molecular characterization of fungal communities in coffee, forest Rubiaceae, and forest non-Rubiaceae leaves. Numbers outside parentheses denote the number of leaves imaged and numbers inside parentheses denote the number of pooled molecular samples from these leaves. Fungal communities were characterized via ITS2 metabarcoding using the fungal-specific primers ITS4 (White et al., 1990 ) and 5.8SR (Vilgalys & Hester, 1990 ). Number of leaf samples by distance by site Sample type Distance from coffee-forest border (m) AJ AN DV EP LZ RB RN VP Total Coffee -100 4(1) 4(1) 5(1) 4(1) 4(1) 3(1) - 4(1) 28(7) -50 4(1) 4(1) 5(1) 4(1) 4(1) 4(1) 8(1) 4(1) 37(8) -25 4(0) 4(1) 5(1) 4(1) 4(1) 4(1) 4(1) 4(1) 33(7) -10 4(1) 4(1) 5(1) 4(1) 4(1) 4(1) 4(1) 4(1) 33(8) Total 16(3) 16(4) 20(4) 16(4) 16(4) 15(4) 16(3) 16(4) 131(30) Forest Rubiaceae 10 12(4) 13(3) 10(2) 15(3) 14(4) 12(3) 8(2) 16(5) 100(26) 25 12(4) 13(4) 15(3) 12(3) 13(4) 20(5) 12(3) 16(5) 113(31) 50 17(4) 12(3) 20(6) 12(3) 11(4) 16(4) 13(4) 17(6) 118(34) 100 12(4) 12(3) 19(5) 13(4) 16(4) 21(4) 12(3) 16(4) 121(31) Total 53(16) 50(13) 64(16) 52(13) 54(16) 69(16) 45(12) 65(20) 452(122) Forest non-Rubiaceae 10 8(2) 11(3) 9(2) 12(2) 12(2) 12(3) 8(0) 13(2) 85(16) 25 11(1) 8(2) 10(2) 12(3) 15(4) 11(3) 8(2) 16(2) 91(19) 50 13(1) 10(3) 5(1) 12(3) 12(3) 12(3) 8(2) 15(2) 87(18) 100 12(0) 7(2) 15(2) 12(3) 12(3) 11(2) 8(2) 15(2) 92(16) Total 44(4) 36(10) 39(7) 48(11) 51(12) 46(11) 32(6) 59(8) 355(69) Statistical analyses Visual assessment of foliar disease symptoms and fungal damage of seedlings We conducted all analyses in R v4.2.3 (R Core Team, 2024). We used a two-step modeling approach to (1) determine whether leaf spot incidence was higher in coffee than in forest leaves, and (2) examine if leaf spot incidence changed with distance into the forest for forest Rubiaceae (native coffee relatives) versus forest non-Rubiaceae hosts. To test whether leaf spot incidence was greater in coffee than in forest leaves, we fitted a binomial generalized linear mixed model (GLMM) with presence/absence of leaf spot as the response, vegetation type (“coffee” vs. “forest”) as a fixed effect, and both site and plant species as random effects to account for site-to-site variation and variable sample sizes of given species. We used the Laplace approximation to ensure stable estimation despite sparse or uneven counts (Rue et al., 2009 ). To determine whether close relatives of coffee (plants within Rubiaceae) exhibited a different pattern of edge-related leaf spot symptoms compared to more distantly related forest hosts, we classified each forest sample as belonging to either a Rubiaceae (close relatives of coffee) or a non-Rubiaceae host. We ran binomial GLMMs predicting leaf spot presence/absence as a function of the fixed effects physical distance from the coffee–forest border and phylogenetic distance of the host species from coffee, plus their interaction, with site as a random effect for Rubiaceae and non-Rubiaceae forest plants. With respect to our seedling fungal damage surveys, we used a mixed linear regression model to explain change in fungal damage at the plot level as a function of physical distance from coffee field, phylogenetic distance from coffee, and their interaction term with forest site as a random effect to account for site variation. Airborne fungi To evaluate whether total airborne fungal propagule load changed with distance from coffee into forest, we pooled monthly samples within sites to obtain an aggregated propagule load at each distance, avoiding potential pseudoreplication at each sampling location across months.We used a negative binomial mixed model on the counts along the transects with distance as a continuous fixed effect and site as a random effect. We used a beta-binomial mixed model with a logit link with the relative abundance of OTUs matched to the fungal order Pleosporales, which was the most abundant and diverse order detected on the coffee leaves (Fig. 2 ; Table S1 ) and includes many coffee pathogens, making it a strong candidate group for investigating coffee-to-forest spillover of fungi. These analyses were conducted using the “lme4” (v. 1.1–31) and “glmmTMB” (v.1.1.12) packages (Bates et al., 2025 ). We selected a beta-binomial approach because it can handle overdispersion and is commonly used to model compositional microbiome data (Martin et al., 2020 ). Foliar fungi To assess if host relatedness to coffee plays a role in the abundance of coffee-associated fungal taxa in the host phyllosphere, we calculated the phylogenetic distance of each of our forest plant species from coffee using methods described by Sternhagen ( 2019 ). Briefly, we acquired NCBI sequences spanning the 18S, ITS1, 5.8S, ITS2 and 26S regions of DNA for each of the plant species used in this study or the closest relatives with relevant sequence data (Table 2; Benson et al., 2018 ). All sequences we used contained at least a complete 5.8S segment. Phylogenetic distance from coffee for each understory plant species was extracted using the Tamura-Nei model with complete deletion via MEGA version 11.0.13 (Kumar et al., 2018 ). The phylogenetic distance of coffee to itself was set at zero. To evaluate if the relative abundance of Pleosporales declined from coffee into forest as with the air samplers, we fitted a beta-binomial mixed model, with the relative abundance of Pleosporales as the response variable, physical distance from the coffee-forest border as a fixed effect, and site as a random effect to partition variance due to sample location. To test whether the relative abundance of Pleosporales on forest leaves was influenced by proximity to the coffee field and the host plant's relatedness to coffee, we took a stratified approach and fitted the above model for forest Rubiaceae leaves and forest non-Rubiaceae separately. Phylogenetic distance of host plants from coffee was not a strong predictor of Pleosporales relative abundance in our Rubiaceae or non-Rubiaceae forest models and was removed from our final models. For final model selection, we compared all possible combinations of our variables of interest and selected the model with the lowest Akaike Information Criterion (AIC) value, following Burnham and Anderson ( 2004 ). Investigating spillover of individual taxa Finally, to determine if individual fungal taxa from the air and leaves attenuated with distance into the forest, we first excluded taxa from the OTU matrix that appeared fewer than five times, as these would not yield a detectable signal across distances. We used generalized joint attribute modeling (GJAM; Clark et al., 2017 ) to test for fungal taxa that declined with distance into the forest interior using data from the forest air samplers (forest-facing at the border, and at 10, 25, 50, 100 m into the forest). We used physical distance from the coffee-forest border as a continuous fixed effect, sampling month as a categorical fixed effect, and site as a random effect. This modeling approach can accommodate multivariate datasets composed of various data types that are dominated by zeros in its probabilistic modeling approach (Taylor-Rodríguez et al., 2017 ). For forest leaf samples, the OTU matrix was too large to run in its entirety, even with the dimension reduction step built into this modeling approach. Consequently, we assessed the forest patterns of the 1,000 most abundant OTUs observed in coffee that also appeared on > 10 pooled leaf samples in the forest (~ 56% of the total reads in the leaf samples), including all other OTUs as a binned “other” category in the model. We ran this full model with physical distance into the forest from the coffee-forest border and phylogenetic distance from coffee of the host plant as continuous fixed effects and site as a random effect. For both GJAM models we used three chains of 5,000 iterations and 500 burn-in iterations to achieve chain convergence. We matched the genera of OTUs to their ecological guild in order to determine the probable ecological role of these taxa and their potential as plant pathogens using the “Primary Lifestyle” designation of the FungalTraits database (Põlme et al., 2020 ). For OTUs that could not be assigned to a genus, we manually assigned a guild designation when possible. Results Leaf spot on coffee and forest understory leaves and seedling fungal damage We found the odds of leaf spot incidence were lower on forest understory plants than on coffee (Fig. 3 A). In forest Rubiaceae, the odds of leaf spot incidence declined with distance into the forest from the coffee–forest border (Fig. 3 B). However, for non-Rubiaceae plants in the forest understory, there was no effect of distance from the border on leaf spot incidence ( β = − 0.06, z = − 0.28, SE = 0.21, P = 0.7790). There was no effect of phylogenetic distance from coffee on incidence of leaf spot in forest Rubiaceae or non-Rubiaceae (β = − 0.18, z = − 0.76, SE = 0.24, P = 0.449; β = 0.53, z = 1.66, SE = 0.32, P = 0.0971; data not shown) and no interaction between distance and phylogenetic distance for either Rubiaceae or non-Rubiaceae (β = 0.18, z = 0.725, SE = 0.25. P = 0.4685; β = − 0.22, z = − 0.98, SE = 0.23, P = 0.3284; data not shown). At the four sites where we surveyed seedling damage, there was a positive interaction between phylogenetic distance from coffee and physical distance from the coffee field in predicting fungal damage at the plot level (β = 0.517 ± 0.202, t = 2.57, P = 0.015; Fig. 4 ), indicating that the negative effect of physical distance is strongest in hosts closely related to coffee and becomes progressively weaker in more distantly related hosts. Airborne fungi from coffee to forest From the 203 air samples we collected, the abundance table contained 2,465,390 reads and 6,838 OTUs. Total fungal read counts in the air increased with distance along the transect into the forest (β = 0.005 ± 0.002 SE, z = 2.26, P = 0.0236; Fig. 5 A). At the coffee-forest border, there were more read counts on average from air samplers facing the forest then those facing the coffee (β = 2.62 ± 1.27 SE, z = 2.07, P = 0.0389). In contrast, the relative abundance of Pleosporales declined significantly with increasing distance into the forest (β = − 0.012 ± 0.003 SE, z = − 3.85, P = 0.0001) whether border-forest samplers were included or only forest samplers (Table S6). Among the fungal OTUs detected in the air, 33 declined in relative abundance from the coffee-forest border into the forest interior (Fig. 6 ). These taxa included saprotrophs (13 OTUs), and potential plant pathogens/endophytes (9 OTUs), two entomopathogens, one epiphyte and one mycoparasite. Seven of the 33 could not be assigned to an ecological guild. Foliar fungi from the coffee-forest border into the forest interior After quality control steps, the OTU abundance table for foliar fungal communities contained 6,539,462 reads and 9,905 OTUs. The relative abundance of Pleosporales declined with physical distance into the forest on Rubiaceae but not non-Rubiaceae (Fig. 7 B). Of the 1,000 most abundant OTUs in coffee (hereafter “coffee-associated” taxa), 836 were observed on > 10 pooled sets of forest understory leaves, making up the majority (52.4%) of the total reads in the dataset. Of these 836 OTUs, 19 (< 3%) declined in relative abundance with increasing distance into the forest (Fig. 8 A; Table S3). These included several taxa whose closest matches were potential plant pathogens (10 OTUs), which may also function as endophytes depending on host context, as well as closest matches to nematophagous (2 OTUs), saprotrophic (2 OTUs), a mycoparasitic (1 OTU), and lichenized (1 OTU) taxa. Among the 836 coffee-associated OTUs found on at least 10 molecular samples of forest understory plants, 93 (~ 11%) exhibited greater relative abundance the more closely related the forest host species was to coffee (Fig. 8 B). Among these taxa, nearly half of those assigned to an ecological guild were potential plant pathogens and endophytes (33 OTUs). The remaining taxa for which it was possible to assign a putative ecological guild included saprotrophs (25 OTUs), epiphytes and lichenized fungi (6 OTUs), mycoparasites (3 OTUs), and one nematophagous OTU (Fig. 8 B; Table S4). An additional 33 OTUs could not be assigned to a guild. For 10 OTUs, there was an interaction of physical distance with phylogenetic distance, such that further from the edge, fungal read counts were less affected by the relatedness of the forest understory host plant to coffee (Fig. 8 C). Discussion We illustrate that coffee can act as a reservoir for foliar fungi, including plant pathogens, increasing disease pressure in adjacent forests for native plants closely related to the crop plant. Several lines of evidence support this: Visual surveys showed leaf spots were much more common on coffee leaves than on understory forest leaves. In forest Rubiaceae, the odds of leaf spot incidence declined with distance from the coffee–forest border. Likewise, fungal damage on forest seedlings decreased sharply with distance from coffee fields in close coffee relatives, but this effect weakened in more distantly related hosts. Pleosporales, the dominant fungal order in coffee fields, were more abundant on forest Rubiaceae closer to the coffee–forest border, mirroring their spatial pattern in airborne propagules. Lastly, among the 1,000 most abundant OTUs in coffee, roughly 10% were more abundant on plants the more closely related they were to coffee. The relative abundance of 19 of these OTUs, including many taxa with known potential as plant pathogens, declined with distance into the forest. This work demonstrates an underappreciated biotic edge effect of agricultural land adjacent to wildland habitats. Previous work on ecological spillover has focused on natural ecosystems as reservoirs of biodiversity that influence adjacent agricultural systems via pollinators (Garibaldi et al., 2011 ), pests (Wisler & Norris, 2005 ), or pathogens moving into crop system (Blitzer et al., 2012 ). Our findings suggest this phenomenon can operate in reverse: intensive crop systems may act as reservoirs of crop-associated fungi, including those with pathogenic potential, that move into adjacent forest and increase disease pressure near habitat edges. This represents a biotic edge effect distinct from more commonly documented effects such as changes in light, humidity, temperature, canopy cover or plant density (Harper et al., 2005 ; Willmer et al., 2022 ). Over time, crop-driven spillover of pathogens could restructure adjacent forest plant communities by selectively disadvantaging plant taxa closely related to coffee. In this way, agricultural landscapes may not merely fragment habitats but actively restructure adjacent plant and fungal communities. The patterns we observed support our hypotheses that coffee-associated foliar fungi alter the foliar fungal communities of adjacent forest plants, particularly those closely related to coffee. For example, leaf spot, which was much frequent in coffee than forest, declined with distance into the forest away from the coffee-forest border only for forest Rubiaceae (Figs. 3 B, 3 C). This pattern parallels the decline with distance into the forest of Pleosporales on Rubiaceae leaves (Fig. 7 B), and in the air (Fig. 5 B). Indeed, the relative abundance of Pleosporales on forest Rubiaceae leaves was negatively associated with phylogenetic distance from coffee (Fig. 7 C), suggesting that these fungi more readily colonize close relatives of the cultivated host. Together, these patterns support a host-specific spillover of foliar fungi from coffee into adjacent forest disproportionately affecting native Rubiaceae plants, facilitated by airborne coffee-associated fungal propagules Importantly, spatial decline in disease was restricted to Rubiaceae; non-Rubiaceae hosts showed no such pattern. If microclimatic variation near the forest edge were the primary driver of elevated disease pressure, we would expect a uniform response across host lineages. Instead, the pattern was specific to Rubiaceae, suggesting a host-specific mechanism consistent with spillover from coffee, rather than changes in humidity, temperature, or light exposure acting as environmental filters (Laurance et al., 2002). Additionally, if microclimatic conditions alone were shaping leaf-associated fungal communities, we would expect a decoupling between airborne and leaf-surface communities (O’Malley, 2008; Wright et al., 2024). Instead, we observed parallel patterns for Pleosporales in air and leaves, suggesting Pleosporales propagule load in the coffee amplifies their abundance on compatible nearby forest hosts. We identified over 100 coffee-associated fungal taxa that appeared to exhibit host affinity for forest plants in the coffee family (Fig. 8 B, 8 C), a pattern consistent with known phylogenetic constraints on pathogen host ranges (Gilbert & Webb, 2007 ). Over 25% of these taxa had pathogenic potential, including members whose closest taxonomic match came from the genera Colletotrichum , Setophoma , Cercospora , and the family Didymellaceae which contains pathogens (e.g. Colletotrichum kahawe, Setophoma terrestris, Cercospora coffeicola, and Phoma costaricensis) known to drive coffee leaf spot epidemics in Neotropical coffee fields (Fig. 8 B; Table S4; Aveskamp et al., 2010 , Deb et al., 2020 ; Aparecido et al., 2024 ; ). Intriguingly, in the subset of these taxa (~ 10%, including 4 Didymellaceae taxa) in which the effect of phylogenetic distance depended on physical distance from the coffee field, the importance of phylogenetic distance diminished with increasing distance into the forest (Fig. 8 C). This dissipating interaction suggests that while evolutionary relationships determine host susceptibility, realized colonization appears contingent on spatial proximity to the coffee field. Broadly, these spillover dynamics could disrupt the density-dependent processes that structure forest understory communities. By increasing pathogen pressure near forest edges in proportion to the density of adjacent coffee plants, agricultural systems may impose a competing, externally driven density-dependent process capable of reshaping forest plant communities in fragmented landscapes. While the forest understory appears to harbor a broader consortium of airborne fungal propagules than coffee (Fig. 5 A), certain crop-associated groups, particularly Pleosporales, appear augmented in coffee and decline in abundance with distance into the forest (Fig. 5 B). This indicates that coffee acts not only as a source of high-density host-specific propagules but also selectively alters the composition of the fungal pool at forest edges. Janzen–Connell theory posits that host-specific enemies reduce survival near conspecific adults, limiting dominance by common species and allowing rarer species to persist, where plants at intermediate densities are most vulnerable to this effect (Connell et al., 1984; Bachelot et al., 2016). Yet the very mechanism that should promote the persistence of rare species closely related to the crop plant—their low density—may render these species most vulnerable at crop–forest borders, where coffee-associated pathogens disproportionately infect close relatives of the crop. Conservation ecologists should explore how targeted management strategies, such as buffer vegetation composed of species distantly related to the crop (e.g., forage grasses between coffee fields and forest), could mitigate pathogen spillover into adjacent natural areas. While the spatial distribution of fungal taxa is consistent with spillover from coffee, this work cannot directly trace the origin of fungal propagules found on or in forest plant tissues. Confirming that forest-colonizing foliar fungi originate in coffee fields requires directly tracking fungal movement via techniques like isotopic labeling or whole-genome sequencing, a critical next step for understanding the role of coffee fields as source populations for foliar fungi in surrounding landscapes. Another important follow-up would be to examine Rubiaceae seedling survival in forest fragments with or without adjacent coffee. Conclusion High-density crop systems such as coffee fields can influence fungal communities in adjacent forests through ecological spillover, disproportionately affecting forest understory plants that are closely related to the crop plant. Our findings suggest that coffee fields serve as reservoirs for coffee-associated plant pathogens more likely to colonize nearby Rubiaceae plants in the forest understory. In the mosaic landscapes of the Neotropics. where agricultural land is in direct proximity to both forest fragments and protected forest areas, crop-mediated spillover of coffee pathogens into forest could alter forest plant community composition by amplifying disease pressure on native Rubiaceae. This presents an underappreciated biotic edge effect that can alter the plant community composition of apparently intact natural ecosystems. It is therefore crucial to promote larger, more contiguous conservation areas, particularly in the highly biodiverse tropics where agricultural encroachment is widespread and increasing. Declarations Acknowledgements We are indebted to the farmers who allowed us access to their coffee farms and forests, without whom none of this work would have been possible. The staff at the Las Cruces Biological Station deserve much gratitude for creating a comfortable and productive place to work. The Organization for Tropical Studies (OTS) provided access to lab space and equipment as well as funding for food and lodging for GE from the Donald & Beverly Stone Fellowships (Fund 507/557). Enrique Castro Fonseca at OTS, and staff at the National Commission for the Management of Biodiversity (CONAGEBIO) in Costa Rica and USDA PPQ in Beltsville, Maryland, USA assisted with securing permits for research and collecting. We also thank Michael Atencio Picado and Erick Barrantes Fallas, who helped deploy and maintain the air samplers. This work was supported by the resources and staff at the University of Minnesota Genomics Center (https://genomics.umn.edu). The bioinformatic pipelines of this study were executed using the Center for Computationally Assisted Science and Technology (CCAST) at North Dakota State University, made possible in part by NSF MRI Award No. 2019077. This material is based upon work supported by the U.S. National Science Foundation Graduate Research Fellowship awarded to JAL (Grant No. 2137101), and NSF CAREER Award 2048131 and Award 2037121 to LAW. References Abarenkov, K., Nilsson, R. H., Larsson, K.-H., Taylor, A. F. S., May, T. W., Frøslev, T. G., Pawlowska, J., Lindahl, B., Põldmaa, K., Truong, C., Vu, D., Hosoya, T., Niskanen, T., Piirmann, T., Ivanov, F., Zirk, A., Peterson, M., Cheeke, T. E., Ishigami, Y., … Kõljalg, U. (2024). The UNITE database for molecular identification and taxonomic communication of fungi and other eukaryotes: Sequences, taxa and classifications reconsidered. Nucleic Acids Research , 52 (D1), D791–D797. https://doi.org/10.1093/nar/gkad1039 Aparecido, L. E. D. O., Lorençone, J. A., Lorençone, P. A., Torsoni, G. B., Lima, R. F. de, Chiquitto, A. G., Saqui, D., Oliveira Júnior, G. G. de, & Rolim, G. de S. (2024). Climate change and Phoma spp. leaf spot of Arabica coffee: A CMIP6 Modeling Approach. Revista Brasileira de Meteorologia , 38 , e38230041. https://doi.org/10.1590/0102-77863810041 Apigo, A., & Oono, R. (2022). Plant abundance, but not plant evolutionary history, shapes patterns of host specificity in foliar fungal endophytes. Ecosphere , 13 (1), e03879. https://doi.org/10.1002/ecs2.3879 Aveskamp, M. M., De Gruyter, J., Woudenberg, J. H. C., Verkley, G. J. M., & Crous, P. W. (2010). Highlights of the Didymellaceae: A polyphasic approach to characterise Phoma and related pleosporalean genera. Studies in Mycology , 65 , 1–60. https://doi.org/10.3114/sim.2010.65.01 Bates, D., Maechler, M., Bolker [aut, B., cre, Walker, S., Christensen, R. H. B., Singmann, H., Dai, B., Scheipl, F., Grothendieck, G., Green, P., Fox, J., Bauer, A., simulate.formula), P. N. K. (shared copyright on, Tanaka, E., & Jagan, M. (2025). lme4: Linear Mixed-Effects Models using “Eigen” and S4 (Version 1.1-36) [Computer software]. https://cran.r-project.org/web/packages/lme4/index.html Benítez-Malvido, J., Lázaro, A., & Ferraz, I. D. K. (2018). Effect of distance to edge and edge interaction on seedling regeneration and biotic damage in tropical rainforest fragments: A long-term experiment. Journal of Ecology , 106 (6), 2204–2217. https://doi.org/10.1111/1365-2745.13003 Benson, D. A., Cavanaugh, M., Clark, K., Karsch-Mizrachi, I., Ostell, J., Pruitt, K. D., & Sayers, E. W. (2018). GenBank. Nucleic Acids Research , 46 (D1), D41–D47. https://doi.org/10.1093/nar/gkx1094 Blitzer, E. J., Dormann, C. F., Holzschuh, A., Klein, A.-M., Rand, T. A., & Tscharntke, T. (2012). Spillover of functionally important organisms between managed and natural habitats. Agriculture, Ecosystems & Environment , 146 (1), 34–43. https://doi.org/10.1016/j.agee.2011.09.005 Burnham, K. P., & Anderson, D. R. (2004). Multimodel Inference: Understanding AIC and BIC in model selection. Sociological Methods & Research , 33 (2), 261–304. https://doi.org/10.1177/0049124104268644 Cadenasso, M. L., Pickett, S. T. A., Weathers, K. C., & Jones, C. G. (2003). A framework for a theory of ecological boundaries. BioScience , 53 (8), 750–758. https://doi.org/10.1641/0006-3568(2003)053[0750:AFFATO]2.0.CO;2 Caudill, S. A., Brokaw, J. N., Doublet, D., & Rice, R. A. (2017). Forest and trees: Shade management, forest proximity and pollinator communities in southern Costa Rica coffee agriculture. Renewable Agriculture and Food Systems , 32 (5), 417–427. https://doi.org/10.1017/S1742170516000351 Clark, J. S., Nemergut, D., Seyednasrollah, B., Turner, P. J., & Zhang, S. (2017). Generalized joint attribute modeling for biodiversity analysis: Median-zero, multivariate, multifarious data. Ecological Monographs , 87 (1), 34–56. https://doi.org/10.1002/ecm.1241 Deb, D., Khan, A., & Dey, N. (2020). Phoma diseases: Epidemiology and control. Plant Pathology , 69 (7), 1203–1217. https://doi.org/10.1111/ppa.13221 Delprete, P. G., & Jardim, J. G. (2012). Systematics, taxonomy and floristics of Brazilian Rubiaceae: An overview about the current status and future challenges. Rodriguésia , 63 , 101–128. https://doi.org/10.1590/S2175-78602012000100009 Garibaldi, L. A., Steffan-Dewenter, I., Kremen, C., Morales, J. M., Bommarco, R., Cunningham, S. A., Carvalheiro, L. G., Chacoff, N. P., Dudenhöffer, J. H., Greenleaf, S. S., Holzschuh, A., Isaacs, R., Krewenka, K., Mandelik, Y., Mayfield, M. M., Morandin, L. A., Potts, S. G., Ricketts, T. H., Szentgyörgyi, H., … Klein, A. M. (2011). Stability of pollination services decreases with isolation from natural areas despite honey bee visits. Ecology Letters , 14 (10), 1062–1072. https://doi.org/10.1111/j.1461-0248.2011.01669.x Gilbert, G. S., & Hubbell, S. P. (1996). Plant diseases and the conservation of tropical forests. BioScience , 46 (2), 98–106. https://doi.org/10.2307/1312812 Gilbert, G. S., & Webb, C. O. (2007). Phylogenetic signal in plant pathogen–host range. Proceedings of the National Academy of Sciences , 104 (12), 4979–4983. https://doi.org/10.1073/pnas.0607968104 Gohl, D. M., Vangay, P., Garbe, J., MacLean, A., Hauge, A., Becker, A., Gould, T. J., Clayton, J. B., Johnson, T. J., Hunter, R., Knights, D., & Beckman, K. B. (2016). Systematic improvement of amplicon marker gene methods for increased accuracy in microbiome studies. Nature Biotechnology , 34 (9), 942–949. https://doi.org/10.1038/nbt.3601 Golan, J. J., & Pringle, A. (2017). Long-distance dispersal of fungi. Microbiology Spectrum , 5 (4). https://doi.org/10.1128/microbiolspec.FUNK-0047-2016 Goossens, D., & Offer, Z. Y. (2000). Wind tunnel and field calibration of six aeolian dust samplers. Atmospheric Environment , 34 (7), 1043–1057. https://doi.org/10.1016/S1352-2310(99)00376-3 Gweon, H. S., Oliver, A., Taylor, J., Booth, T., Gibbs, M., Read, D. S., Griffiths, R. I., & Schonrogge, K. (2015). PIPITS: An automated pipeline for analyses of fungal internal transcribed spacer sequences from the Illumina sequencing platform. Methods in Ecology and Evolution , 6 (8), 973–980. https://doi.org/10.1111/2041-210X.12399 Harper, K. A., Macdonald, S. E., Burton, P. J., Chen, J., Brosofske, K. D., Saunders, S. C., Euskirchen, E. S., Roberts, D., Jaiteh, M. S., & Esseen, P.-A. (2005). Edge influence on forest structure and composition in fragmented landscapes. Conservation Biology , 19 (3), 768–782. https://doi.org/10.1111/j.1523-1739.2005.00045.x Harvey, C. A., Pritts, A. A., Zwetsloot, M. J., Jansen, K., Pulleman, M. M., Armbrecht, I., Avelino, J., Barrera, J. F., Bunn, C., García, J. H., Isaza, C., Munoz-Ucros, J., Pérez-Alemán, C. J., Rahn, E., Robiglio, V., Somarriba, E., & Valencia, V. (2021). Transformation of coffee-growing landscapes across Latin America. A review. Agronomy for Sustainable Development , 41 (5), 62. https://doi.org/10.1007/s13593-021-00712-0 Holdridge, L. R., & Grenke, W. C. et al. (1971). Forest environments in tropical life zones: A pilot study. Forest Environments in Tropical Life Zones: A Pilot Study. https://www.cabdirect.org/cabdirect/abstract/19716605728 Johnson, B. L., & Haddad, N. M. (2011). Edge effects, not connectivity, determine the incidence and development of a foliar fungal plant disease. Ecology , 92 (8), 1551–1558. https://doi.org/10.1890/10-1072.1 Kumar, S., Stecher, G., Li, M., Knyaz, C., & Tamura, K. (2018). MEGA X: Molecular Evolutionary genetics analysis across computing platforms. Molecular Biology and Evolution , 35 (6), 1547–1549. https://doi.org/10.1093/molbev/msy096 Laurance, W. F., Nascimento, H. E. M., Laurance, S. G., Andrade, A., Ewers, R. M., Harms, K. E., Luizão, R. C. C., & Ribeiro, J. E. (2007). Habitat fragmentation, variable edge effects, and the landscape-divergence hypothesis. PLOS ONE , 2 (10), e1017. https://doi.org/10.1371/journal.pone.0001017 Ma, J., Li, J., Wu, W., & Liu, J. (2023). Global forest fragmentation change from 2000 to 2020. Nature Communications , 14 (1), 3752. https://doi.org/10.1038/s41467-023-39221-x Martin, B. D., Witten, D., & Willis, A. D. (2020). Modeling microbial abundances and dysbiosis with beta-binomial regression. The Annals of Applied Statistics , 14 (1), 94–115. https://doi.org/10.1214/19-AOAS1283 McMurdie, P. J., & Holmes, S. (2014). Waste not, want not: Why rarefying microbiome data is inadmissible. PLOS Computational Biology , 10 (4), e1003531. https://doi.org/10.1371/journal.pcbi.1003531 Pajares-Murgó, M., Garrido, J. L., Perea, A. J., López-García, Á., Bastida, J. M., & Alcántara, J. M. (2024). Mutualistic and antagonistic phyllosphere fungi contribute to plant recruitment in natural communities. Journal of Ecology , 113 (6). https://doi.org/10.1111/1365-2745.14311 Pendrill, F., Gardner, T. A., Meyfroidt, P., Persson, U. M., Adams, J., Azevedo, T., Bastos Lima, M. G., Baumann, M., Curtis, P. G., De Sy, V., Garrett, R., Godar, J., Goldman, E. D., Hansen, M. C., Heilmayr, R., Herold, M., Kuemmerle, T., Lathuillière, M. J., Ribeiro, V., … West, C. (2022). Disentangling the numbers behind agriculture-driven tropical deforestation. Science , 377 (6611), eabm9267. https://doi.org/10.1126/science.abm9267 Põlme, S., Abarenkov, K., Henrik Nilsson, R., Lindahl, B. D., Clemmensen, K. E., Kauserud, H., Nguyen, N., Kjøller, R., Bates, S. T., Baldrian, P., Frøslev, T. G., Adojaan, K., Vizzini, A., Suija, A., Pfister, D., Baral, H.-O., Järv, H., Madrid, H., Nordén, J., … Tedersoo, L. (2020). FungalTraits: A user-friendly traits database of fungi and fungus-like stramenopiles. Fungal Diversity , 105 (1), 1–16. https://doi.org/10.1007/s13225-020-00466-2 Razafimandimbison, S. G., & Rydin, C. (2024). Phylogeny and classification of the coffee family (Rubiaceae, Gentianales): Overview and outlook. TAXON , 73 (3), 673–717. https://doi.org/10.1002/tax.13167 Reis Medeiros, H., Martello, F., Metzger, J. P., Harper, K. A., Mengual, X., Righi, C. A., & Ribeiro, M. C. (2022). Landscape composition regulates the spillover of beneficial insects between forest remnants and adjacent coffee plantations. Perspectives in Ecology and Conservation , 20 (2), 111–116. https://doi.org/10.1016/j.pecon.2021.11.003 Rice, R. A. (1999). A place unbecoming: The coffee farm of northern Latin America. Geographical Review , 89 (4), 554–579. https://doi.org/10.1111/j.1931-0846.1999.tb00234.x Rue, H., Martino, S., & Chopin, N. (2009). Approximate Bayesian inference for latent Gaussian models by using integrated nested laplace approximations. Journal of the Royal Statistical Society Series B: Statistical Methodology , 71 (2), 319–392. https://doi.org/10.1111/j.1467-9868.2008.00700.x Spear, J. E., Grijalva, E. K., Michaels, J. S., & Parker, S. S. (2018). Ecological spillover dynamics of organisms from urban to natural landscapes. Journal of Urban Ecology , 4 (1), juy008. https://doi.org/10.1093/jue/juy008 Sternhagen, E. C. (2019). Influences on Fungal Community Composition in Roots and Soil of Coffee and Other Rubiaceae in Costa Rica [North Dakota State University]. https://hdl.handle.net/10365/29307 Sternhagen, E. C., Black, K. L., Hartmann, E. D. L., Shivega, W. G., Johnson, P. G., McGlynn, R. D., Schmaltz, L. C., Asheim Keller, R. J., Vink, S. N., & Aldrich-Wolfe, L. (2020). Contrasting patterns of functional diversity in Coffee Root Fungal Communities associated with organic and conventionally managed fields. Applied and Environmental Microbiology , 86 (11), e00052-20. https://doi.org/10.1128/AEM.00052-20 Taylor-Rodríguez, D., Kaufeld, K., Schliep, E. M., Clark, J. S., & Gelfand, A. E. (2017). Joint species distribution modeling: Dimension reduction using Dirichlet processes. Bayesian Analysis , 12 (4), 939–967. https://doi.org/10.1214/16-BA1031 Vilgalys, R., & Hester, M. (1990). Rapid genetic identification and mapping of enzymatically amplified ribosomal DNA from several Cryptococcus species. Journal of Bacteriology , 172 (8), 4238–4246. https://doi.org/10.1128/jb.172.8.4238-4246.1990 White, T. J., Bruns, T., & Lee, S. (1990). Analysis of phylogenetic relationships by amplification and direct sequencing of ribosomal RNA genes. A Guide to Methods and Applications . Willmer, J. N. G., Püttker, T., & Prevedello, J. A. (2022). Global impacts of edge effects on species richness. Biological Conservation , 272 , 109654. https://doi.org/10.1016/j.biocon.2022.109654 Wisler, G. C., & Norris, R. F. (2005). Interactions between weeds and cultivated plants as related to management of plant pathogens. Weed Science , 53 (6), 914–917. https://doi.org/10.1614/WS-04-051R.1 Zahawi, R. A., Duran, G., & Kormann, U. (2015). Sixty-seven years of land-use change in southern Costa Rica. PLOS ONE , 10 (11), e0143554. https://doi.org/10.1371/journal.pone.0143554 Table 2 Table 2 is not available with this version. Additional Declarations There is NO Competing Interest. Supplementary Files SpilloverCRsupplementnature.docx Supplementary Materials Cite Share Download PDF Status: Under Review Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7781221","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Biological Sciences - Article","associatedPublications":[],"authors":[{"id":526501337,"identity":"9259584e-f0b8-4fc5-b9fd-fb7becd04217","order_by":0,"name":"Jeffrey Lackmann","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA1UlEQVRIiWNgGAWjYBACPgYGZjCDXwJEGoDIBPxa2GBaJGeQrMXgBlyMkBb2w4+NbtTcyTO+3fxM4keBBQM/e44Bfi08acbJOceeFZvdOWYm2QN0mGTPGwJaGHKYD+ewHU7cdiPB7AYPUIvBDUK28L8Bavl3OHHzjPRvN/8AtdgT1CKRw5yc23Y4cYNEjtltsC0SBLU8MzbO7TucOONGTvlvGQMJHokzzwrwauHnT34snfPtcGL/jPTNhm/+1MnxtydvwKsFA/CQpnwUjIJRMApGAVYAAEOCQrEEBgFiAAAAAElFTkSuQmCC","orcid":"https://orcid.org/0000-0002-7541-6487","institution":"North Dakota State University","correspondingAuthor":true,"prefix":"","firstName":"Jeffrey","middleName":"","lastName":"Lackmann","suffix":""},{"id":526501338,"identity":"7bb2bb7e-4cdd-4df0-9ff7-3e5ff250f6d2","order_by":1,"name":"Benedicte Bachelot","email":"","orcid":"https://orcid.org/0000-0003-3348-9757","institution":"Oklahoma State University","correspondingAuthor":false,"prefix":"","firstName":"Benedicte","middleName":"","lastName":"Bachelot","suffix":""},{"id":526501339,"identity":"fdc8ea81-3c19-4d65-bd6d-4ebfe0dcfb9b","order_by":2,"name":"Catherine Lindell","email":"","orcid":"","institution":"Michigan State University","correspondingAuthor":false,"prefix":"","firstName":"Catherine","middleName":"","lastName":"Lindell","suffix":""},{"id":526501340,"identity":"fd660a06-ee4d-4cb4-baa6-822d421cd49f","order_by":3,"name":"Elena Prado-Ragan","email":"","orcid":"","institution":"North Dakota State University","correspondingAuthor":false,"prefix":"","firstName":"Elena","middleName":"","lastName":"Prado-Ragan","suffix":""},{"id":526501341,"identity":"073a9841-7d02-4974-a194-8c2de5984515","order_by":4,"name":"Priscila Chaverri","email":"","orcid":"","institution":"Bowie State University","correspondingAuthor":false,"prefix":"","firstName":"Priscila","middleName":"","lastName":"Chaverri","suffix":""},{"id":526501342,"identity":"2544e9f6-3950-48df-a452-8d6b5a030631","order_by":5,"name":"Efraín Escudero-Leyva","email":"","orcid":"","institution":"Universidad de Costa Rica","correspondingAuthor":false,"prefix":"","firstName":"Efraín","middleName":"","lastName":"Escudero-Leyva","suffix":""},{"id":526501343,"identity":"8404c282-69d8-4a39-8e3f-8a5d2d438bb4","order_by":6,"name":"Gina Errico","email":"","orcid":"","institution":"Oklahoma State University","correspondingAuthor":false,"prefix":"","firstName":"Gina","middleName":"","lastName":"Errico","suffix":""},{"id":526501344,"identity":"75d37556-4b16-45e3-835f-c55dffbc6f80","order_by":7,"name":"Megan Orr","email":"","orcid":"","institution":"North Dakota State University","correspondingAuthor":false,"prefix":"","firstName":"Megan","middleName":"","lastName":"Orr","suffix":""},{"id":526501345,"identity":"9e920d24-9983-4b87-bf35-ee2f9ceb274d","order_by":8,"name":"Federico Oviedo Brenes","email":"","orcid":"","institution":"Ecos del Bosque","correspondingAuthor":false,"prefix":"","firstName":"Federico","middleName":"Oviedo","lastName":"Brenes","suffix":""},{"id":526501346,"identity":"b3eb1d09-d699-4187-a1d6-25f87701af04","order_by":9,"name":"Jeisson Figueroa","email":"","orcid":"","institution":"Organization for Tropical Studies","correspondingAuthor":false,"prefix":"","firstName":"Jeisson","middleName":"","lastName":"Figueroa","suffix":""},{"id":526501347,"identity":"cd7d2d71-5ec7-4660-98ee-83ff3dc69ea3","order_by":10,"name":"Laura Aldrich-Wolfe","email":"","orcid":"https://orcid.org/0000-0001-9758-701X","institution":"North Dakota State University","correspondingAuthor":false,"prefix":"","firstName":"Laura","middleName":"","lastName":"Aldrich-Wolfe","suffix":""}],"badges":[],"createdAt":"2025-10-04 17:15:08","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7781221/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7781221/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":94612552,"identity":"5c89803d-7d33-410a-8d6d-f05840eec3c8","added_by":"auto","created_at":"2025-10-29 02:10:54","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":877159,"visible":true,"origin":"","legend":"\u003cp\u003eCoffee–forest transects were established at eight sites (at one of these, the transect extended only 50 m into coffee due to field size). Rotating passive air samplers were deployed to capture airborne fungal propagules at 50 m from the coffee-forest border in the coffee and at 10, 25, 50, and 100 m into the forest, and fixed-position passive air samplers were located at the border, one facing coffee and the other forest, to sample propagules moving from one habitat to the other. Single leaves were collected from 4-5 coffee plants at 10, 25, 50 m for all sites and 100 m into coffee at seven sites, and from 2-6 individuals of three plant species in the Rubiaceae and two plant species in non-Rubiaceae families at 10, 25, 50 and 100 m into the forest. If sufficient individuals could not be located, additional species were sampled. Leaves were photographed, screened for pathogen damage, and a sample was removed for DNA extraction.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7781221/v1/305029076f8e82bb21b9f519.png"},{"id":94608898,"identity":"62d9c07e-15be-44ca-b684-b0fad2cbba35","added_by":"auto","created_at":"2025-10-28 23:52:29","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1331745,"visible":true,"origin":"","legend":"\u003cp\u003eRelative abundance (\u003cstrong\u003eA\u003c/strong\u003e) and diversity of OTUs (\u003cstrong\u003eB\u003c/strong\u003e) on coffee leaves of the ten most abundant fungal orders for which a taxonomic designation could be made. Reads assigned to Pleosporales had nearly three times higher relative abundance than the next most abundant order, and they exhibited the greatest richness, making the order a strong candidate coffee-associated group to assess for spillover into adjacent forest.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7781221/v1/9ef27a94eff01a63502e8b04.png"},{"id":94608902,"identity":"a70d85b2-87b9-4830-8a72-b9a239d65bc1","added_by":"auto","created_at":"2025-10-28 23:52:29","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1010607,"visible":true,"origin":"","legend":"\u003cp\u003eThe proportion of leaves with leaf spot at each sampling distance \u003cstrong\u003eA)\u003c/strong\u003e for \u003cem\u003eCoffea arabica \u003c/em\u003ein eight coffee fields, \u003cstrong\u003eB)\u003c/strong\u003e for understory forest plants in the coffee family, Rubiaceae, \u003cstrong\u003eC) \u003c/strong\u003efor forest understory plants in non-Rubiaceae families. Sample sizes varied by distance (Table 1)\u003cstrong\u003e. \u003c/strong\u003eValues are means \u003cu\u003e+\u003c/u\u003e SE. \u003cem\u003eY\u003c/em\u003e-axis scales differ for coffee vs. forest data. Coffee leaves were significantly more likely to exhibit leaf spot than forest leaves (β = –2.46, SE = 0.62, z = –3.94, \u003cem\u003eP\u003c/em\u003e \u0026lt; 0.0001). Leaf spot incidence declined with distance into the forest for native Rubiaceae (β = –0.59, SE = 0.28, \u003cem\u003ez\u003c/em\u003e = –2.12, \u003cem\u003eP\u003c/em\u003e = 0.0342) but not non-Rubiaceae (β = –0.06, SE = 0.21, \u003cem\u003ez\u003c/em\u003e = –0.28, \u003cem\u003eP\u003c/em\u003e = 0.7790). \u003csup\u003e\u0026nbsp;\u003c/sup\u003e\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7781221/v1/34492d26b620bba2bb884743.png"},{"id":94608901,"identity":"59d8850c-caf5-405f-9b06-b24c8a9e125f","added_by":"auto","created_at":"2025-10-28 23:52:29","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":704075,"visible":true,"origin":"","legend":"\u003cp\u003eRelationship between distance from the coffee–forest border and pathogen damage on forest seedlings surveyed in 1 m² plots (n = 10) spaced every 10 m from 5–105 m inside the forest at four of the eight sites. Dots are colored from red (closely related to coffee) to blue (distantly related) based on the average phylogenetic distance to coffee of seedlings in each plot. Pathogen damage was quantified as the mean percentage of leaf area affected per plot. Solid lines show model-predicted relationships from a mixed-effects model with site as a random effect. Damage declined with distance from the border in close relatives of coffee, but this decline weakened as phylogenetic distance increased (\u003cem\u003eβ\u003c/em\u003e = 0.517 ± 0.202 SE, \u003cem\u003et\u003c/em\u003e = 2.57, \u003cem\u003eP\u003c/em\u003e = 0.015).\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7781221/v1/a462d0fde8d0350f6597b953.png"},{"id":94612708,"identity":"2f961c58-ebbf-4e11-a58e-23b8aaea3826","added_by":"auto","created_at":"2025-10-29 02:11:31","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":695812,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003e(A)\u003c/strong\u003e The total reads (log scale) of all airborne fungi and (\u003cstrong\u003eB)\u003c/strong\u003e relative abundance of Pleosporales at each sampling distance along the transect pooled across all sampling months (\u003cem\u003en\u003c/em\u003e = 8). Negative values indicate distances into the coffee fields.\u003cem\u003e Y\u003c/em\u003e-axes differ. Total read count increased, while the relative abundance of Pleosporales declined, with distance into the forest (β = 0.005 ± 0.002 SE, \u003cem\u003ez\u003c/em\u003e= 2.26, \u003cem\u003eP\u003c/em\u003e = 0.0236; β = –0.01± 0.003 SE, \u003cem\u003ez\u003c/em\u003e = –3.85, \u003cem\u003eP\u003c/em\u003e= 0.0001).\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-7781221/v1/f94375229630da733bc224ca.png"},{"id":94608903,"identity":"2aa2ff31-9ead-49d8-bd25-0349ea5e0642","added_by":"auto","created_at":"2025-10-28 23:52:30","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":850391,"visible":true,"origin":"","legend":"\u003cp\u003eAirborne OTUs that declined with distance into the forest according to our generalized joint attribute model (GJAM; Clark \u003cem\u003eet al., \u003c/em\u003e2017). Values are mean effect estimates from the model \u003cu\u003e+\u003c/u\u003e SE \u0026nbsp;(Table S2). Airborne samples were collected monthly for four months (for six of the eight sites; two sites were sampled monthly for three and two months, repsectively) using passive air samplers at 0 (forest-facing), 10, 25, 50, and 100 m into the forest interior (\u003cem\u003en\u003c/em\u003e = 29). The size of the negative effect estimate indicates the magnitude of the decline of the taxon with distance into the forest from the edge. Ecological guilds were assigned using FungalTraits (Põlme, \u003cem\u003eet al., \u003c/em\u003e2020) in conjunction with manual screening.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-7781221/v1/ccb902ef32afcbae91b8d101.png"},{"id":94608905,"identity":"2afb92bc-bf9d-43ac-bee6-3e7c4942aa87","added_by":"auto","created_at":"2025-10-28 23:52:30","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":868202,"visible":true,"origin":"","legend":"\u003cp\u003eThe relative abundance of Pleosporales along the transect plotted with results from our beta-binomial GLMM at each sampling distance for (\u003cstrong\u003eA\u003c/strong\u003e) all sampled leaves (negative values denote distance into coffee from coffee-forest edge; Sample sizes vary; see Table 1) \u003cstrong\u003e(B)\u003c/strong\u003e forest Rubiaceae leaves and \u003cstrong\u003e(C)\u003c/strong\u003e forest non-Rubiaceae leaves\u003cstrong\u003e. \u003c/strong\u003eThe relative abundance of Pleosporales decreased with distance along the transect in our model of all leaves (β = -0.01 ± 0.003 SE, \u003cem\u003ez\u003c/em\u003e = -3.85, \u003cem\u003eP\u003c/em\u003e = 0.0001) and with distance from the coffee-forest border in forest Rubiaceae leaves (β = –0.004 ± 0.001 SE, \u003cem\u003ez\u003c/em\u003e = –3.38, \u003cem\u003eP \u003c/em\u003e= 0.0007). No such spatial relationship was observed in non-Rubiaceae forest plants (β = –0.001 ± 0.002 SE, \u003cem\u003ez\u003c/em\u003e = –0.69, \u003cem\u003eP\u003c/em\u003e = 0.493).\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-7781221/v1/6bdff158ec36e2d22f5f1f69.png"},{"id":94612481,"identity":"d5110988-b70a-4326-b2d0-d191e40e7dfd","added_by":"auto","created_at":"2025-10-29 02:10:35","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":683530,"visible":true,"origin":"","legend":"\u003cp\u003eAll leaf fungal OTUs that declined with (\u003cstrong\u003eA)\u003c/strong\u003e distance from the coffee border into the forest, (\u003cstrong\u003eB\u003c/strong\u003e) phylogenetic distance from coffee (Table S3 \u0026amp; S4), and (\u003cstrong\u003eC\u003c/strong\u003e) an interaction of these two fixed effects according to our generalized joint attribute model (GJAM; Clark et al., 2017). Values are means \u003cu\u003e+\u003c/u\u003e SE. For sample sizes see Table 1. Ecological guilds were assigned using FungalTraits (Põlme et al.,\u003cem\u003e \u003c/em\u003e2020) in conjunction with manual screening of existing literature.\u003c/p\u003e","description":"","filename":"8.png","url":"https://assets-eu.researchsquare.com/files/rs-7781221/v1/105976553f0645d23dc6e9e8.png"},{"id":94640013,"identity":"c2258839-91a8-479b-8d5a-6a0a43884976","added_by":"auto","created_at":"2025-10-29 07:47:51","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":8802748,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7781221/v1/f6e2ac5f-1787-4e0c-9c15-2b61bd5701d3.pdf"},{"id":94608900,"identity":"5bbf807e-0362-4fe8-b85c-b3013d9e50bd","added_by":"auto","created_at":"2025-10-28 23:52:29","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":602599,"visible":true,"origin":"","legend":"Supplementary Materials","description":"","filename":"SpilloverCRsupplementnature.docx","url":"https://assets-eu.researchsquare.com/files/rs-7781221/v1/44b3e5a0900cda647269bf84.docx"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Ecological spillover of coffee-associated foliar fungi from field to forest as a biotic edge effect in the Neotropics","fulltext":[{"header":"Introduction","content":"\u003cp\u003eFoliar fungi inhabit leaf surfaces and interiors, exhibit high dispersal potential (Golan \u0026amp; Pringle, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2017\u003c/span\u003e), varying degrees of host specificity (Apigo \u0026amp; Oono, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), and distinct functional roles as both pathogens and mutualists capable of mediating plant community structure (Pajares-Murg\u0026oacute; et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). The widespread conversion of natural ecosystems to agriculture creates mosaic landscapes where managed land directly borders wildlands (Pendrill et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Ma et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). Edges between managed and natural communities create ecological interfaces where foliar fungi from both communities encounter each other. These fungi and their associated functions in the ecosystem can move across these boundaries in a phenomenon known as ecological spillover (Blitzer et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Spear et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eEcotones between agricultural and natural systems are often characterized by pronounced edge effects, where environmental conditions, species composition, and biotic interactions differ markedly from the ecosystem interiors (Rice, \u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Cadenasso et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2003\u003c/span\u003e). In tropical forest landscapes fragmented by agriculture, edges are known to alter microclimate, species turnover, and ecological processes (Ben\u0026iacute;tez-Malvido et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2018\u003c/span\u003e), including the spread of fungal pathogens (Johnson \u0026amp; Haddad, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2011\u003c/span\u003e). High-density monocultures like coffee plantations generate steep biotic gradients at their borders, where elevated abundances of host plants and associated pathogens may increase disease pressure in adjacent natural communities (Laurance et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). While many studies have focused on natural systems as reservoirs of organisms that affect nearby agricultural systems, much less is known about the extent to which organisms from agriculture disperse into and modify adjacent natural systems (Blitzer et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Reis Medeiros et al., \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Many foliar fungi associate closely with managed systems and may accumulate at high enough densities to spill into adjacent wildlands (Gilbert \u0026amp; Hubbell, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e1996\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn forests, Janzen\u0026ndash;Connell effects\u0026mdash;driven by increasing enemy pressure at higher conspecific densities\u0026mdash;limit dominant species and promote the persistence of rare species (Janzen, 1970; Connell, 1971; Liu et al., 2012; Bayandala et al., 2017). These same density-dependent processes can also drive disease outbreaks in managed crop systems when conditions favor host-associated pathogens (Hopkins et al., 2020; Pf\u0026auml;ffle et al., 2015; Wang et al., 2021). Crops grown at high density can amplify populations of pathogens, pests, and natural enemies, increasing the potential for ecological spillover into adjacent natural systems. The likelihood of spillover also depends on host relatedness: foliar fungi are more likely to establish in nearby habitats when they encounter compatible hosts (Gilbert \u0026amp; Webb, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). While some fungi (e.g., \u003cem\u003eXylaria\u003c/em\u003e spp.) have broad host ranges, others (e.g., \u003cem\u003eColletotrichum\u003c/em\u003e spp.) specialize on particular plant lineages (Arnold \u0026amp; Lutzoni, 2007; Qian et al., 2018; Yang et al., 2023; Guo et al., 2024). Thus, plants in natural systems closely related to a crop species should be more susceptible to colonization by crop-associated fungi than more distantly related neighbors (Parker et al., 2015; Kembel \u0026amp; Mueller, 2014; Chen et al., 2022). If these fungi influence host fitness, they could alter plant community composition over time, favoring close relatives of the crop if they act as mutualists or selecting against them if pathogenic (Arnold et al., 2003).\u003c/p\u003e\u003cp\u003eThe foliar fungal communities of coffee plantations in Central America provide an ecologically and economically relevant tropical system to assess whether ecological spillover of crop-associated fungi occurs and whether nearby close relatives of the host plant are more likely to be colonized than more distantly related plant species. Coffee (\u003cem\u003eCoffea arabica\u003c/em\u003e L., Family Rubiaceae), as one of the most economically important crops in the tropics, plays a central role in shaping both rural livelihoods and land use (Harvey et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). These plantations are often managed as high-density monocultures in close proximity to remnant forest (Caudill et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). The Rubiaceae family includes many native forest understory species in Central America (Delprete \u0026amp; Jardim, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Razafimandimbison \u0026amp; Rydin, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This physical proximity to forests and phylogenetic overlap with many forest plants creates conditions where ecological spillover is likely to occur.\u003c/p\u003e\u003cp\u003eWe asked whether coffee plantations act as reservoirs for foliar fungi that can affect forest understory plants. We hypothesized that if coffee serves as a reservoir, we would detect coffee-associated fungi at higher abundances in the air and in the leaves of forest plants near the coffee-forest border compared to plants located further into the forest. Second, we hypothesized that if host relatedness affects colonization, coffee-associated fungi should occur with greater frequency on forest plants the more closely related they are to coffee.\u003c/p\u003e"},{"header":"Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eStudy sites\u003c/h2\u003e\u003cp\u003eFieldwork was conducted March to August 2021 at eight sites in Coto Brus, Puntarenas Province, Costa Rica, near the Las Cruces Biological Station (8\u0026deg;47\u0026prime;7\u0026Prime;N, 82\u0026deg;57\u0026prime;32\u0026Prime;W), in a tropical premontane ecosystem at an elevational range of 800\u0026ndash;1133 m asl with mean annual rainfall of 4000 mm and mean annual temperature of 21\u0026deg;C (Holdridge \u0026amp; Grenke, 1971; Zahawi et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). All sites contained privately owned coffee fields ranging in age from 10 to 67 yrs directly adjacent to primary forest\u0026thinsp;\u0026gt;\u0026thinsp;10 ha. We established a 200 m transect extending 100 m into both the coffee field and the adjacent forest at seven sites ensuring that the 100 m mark was at least 100 m from any other edge of the habitat (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA). At the eighth site (RN), the transect extended only 50 m into the coffee to avoid sampling within 50 m of the opposite edge of the field.\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eLeaf collection for foliar fungi\u003c/h3\u003e\n\u003cp\u003eTo assess the spatial patterns of foliar fungi in coffee fields and the adjacent forest understory, using flame-sterilized scissors we removed a leaf from the second fully expanded leaf pair (counting from the branch tip) from four randomly selected coffee plants within a 5 m radius at 10, 25, 50, and 100 m away from the coffee-forest border at seven sites, while doubling the samples at 50 m in the eighth site, to keep the number of coffee leaf samples consistent across the sites (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). We repeated this sampling process for understory forest plants at 10, 25, 50, and 100 m into the forest, sampling in a 5 m radius from the distance point along transect. To investigate whether forest plants shared more foliar fungi with coffee when closely than more distantly related, we sampled leaves from 2\u0026ndash;6 individuals of each of five plant species (three Rubiaceae, two non-Rubiaceae) at each distance (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e; Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). If at least three individuals of a given species could not be located at a given distance, additional species within the relevant sample category (Rubiaceae or non-Rubiaceae) were sampled. We sampled the same forest plant species across distances and sites whenever possible but, because plant species diversity was so high, we often had to sample different plant species at different distances and sites (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). In total, we sampled 131 coffee leaves (one leaf sample was lost prior to image analysis) and 807 forest leaves (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e) spanning 26 species and eight families (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) from June 1-July 7th, 2021. Representative material from each forest plant species was collected, identified, and deposited in the herbaria at Las Cruces Biological Station and the National Herbarium of Costa Rica.\u003c/p\u003e\n\u003ch3\u003eLeaf imaging for damage morphotyping \u0026 seedling damage survey\u003c/h3\u003e\n\u003cp\u003eThe same day that leaves were collected, they were transported to Las Cruces Biological Station in sterilized coin envelopes at ambient air temperature in sealed plastic bags partially filled with desiccant. At the field station, we photographed each leaf against a uniform background. We manually screened each image for the incidence of disease symptoms including leaf spot, powdery mildew, downy mildew, amorphous chlorosis, and amorphous necrosis. After leaves were photographed, sections of each leaf were haphazardly selected and 800 mg (fresh weight) was removed with a flame-sterilized scalpel, avoiding the midvein. Samples were dried in open 2-mL microcentrifuge tubes placed above silica gel in air-tight containers for ~\u0026thinsp;12 hours and transported to the Aldrich-Wolfe lab (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Samples were lyophilized for 48 h at -80\u0026deg;C using a FreeZone 6 Plus freeze dryer (Labconco Corporation, Kansas City, MO, USA) and subsequently stored at ambient temperature.\u003c/p\u003e\u003cp\u003eIn late May 2022, we established ten 1 \u0026times; 1 m plots over the existing seedling community in a subset of four of the eight forest sites, starting at 5 m from the forest edge and proceeding every 10 m to 105 m into the forest. We individually tagged all woody plant species in the plot, identified them to species, and scored them for leaf fungal damage (%). We assessed the amount of leaf fungal damage as a visual proportion of total leaf damage to undamaged leaf for the entire seedling. Foliar damage was re-measured bi-weekly and seedling mortality status was assessed weekly for eight weeks. If more than 20 individuals of any species were present in a given plot, all were counted but only a random 20 were tagged. Using these data, we calculated total amount of foliar damage at the plot level averaging across seedlings. We also calculated average phylogenetic distance from coffee at the plot level using the R package V.PhyloMaker2.\u003c/p\u003e\n\u003ch3\u003eSampling airborne fungi\u003c/h3\u003e\n\u003cp\u003eTo investigate airborne propagule pressure, Modified Wilson and Cook Samplers (hereafter \u0026ldquo;air samplers\u0026rdquo;; Goossens \u0026amp; Offer, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2000\u003c/span\u003e) were constructed at Las Cruces Biological Station. These wind-driven devices consist of a free-moving mast-mounted pole and aluminum sail that rotates the device with the wind. Each 1.36-m high air sampler was equipped with three wide-mouth 32-oz (946-mL) plastic jars (ULINE, #S-18072, Pleasant Prairie, Wisconsin, USA) positioned at 30, 80, and 130 cm above the soil surface to ensure sampling of air at heights in the typical range of understory shrub species. We fitted each jar with modified lids containing intake and outtake tubes for airflow oriented in relation to the sail such that the intake tube faced into the wind to capture airborne particles (Fig. \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). Prior to field use, collection jars and lids were sterilized in 10% bleach (\u003cem\u003ev/v\u003c/em\u003e) for 30 min and dried for ~\u0026thinsp;45 min at 60\u0026deg;C to evaporate any remaining bleach solution. After drying, the modified lids were sealed onto the jars with plastic wrap covering the intake and outtake tubes until placed in the field.\u003c/p\u003e\u003cp\u003eTo sample airborne fungi, we placed one air sampler each at 10, 25, 50, and 100 m from the coffee-forest border in the forest understory and at 50 m from the coffee-forest border in the coffee field (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). To compare which propagules were moving from coffee to forest and vice versa, we also placed two air samplers at the coffee-forest border on fixed poles that could not rotate, one facing the coffee and the other facing the forest (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Each month, deployed jars were replaced with bleach-sterilized jars in the field, and the deployed jars were sealed, transported to Las Cruces Biological Station, refrigerated, and processed within 24 h.\u003c/p\u003e\u003cp\u003eWe rinsed the contents of the three collection jars from each air sampler with 50 mL of ultrapure water onto a 0.45 \u0026micro;m GN-6 nitrocellulose membrane (Metricel, St. Louis, MO, USA) on a B\u0026uuml;chner funnel attached to a Welch WOB-L Fluid Aspiration Vacuum Pump (Chemtech Scientific, Tampa, FL, USA) to collect particulates including fungal propagules, which are generally larger than 1 \u0026micro;m (Golan \u0026amp; Pringle, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2017\u003c/span\u003e). This resulted in seven samples for each site each month, one for each air sampler, containing the airborne fungi accumulated over the course of the preceding month. Samples were collected at six of the sites in April 2021, at seven of the sites in May 2021, and at all eight sites in June and July 2021, resulting in a total of 203 airborne samples. We included negative controls by processing as above an unused collection jar sterilized and dried identically to those deployed in the field each collection day to determine if any fungi persisted during sterilization or were introduced during filtration. We placed each airborne sample in an individual, sterile 5-mL tube from the DNeasy\u0026reg; PowerWater Kit (Qiagen, Germantown MD, USA). We dried these samples by puncturing tube lids with a flame-sterilized needle, placing them inside a sealed plastic bag filled with silica gel to dry for 24 h at 4\u0026deg;C. Tubes were then sealed with new lids and stored at -20\u0026deg;C. Tubes were shipped frozen on dry ice to North Dakota State University and stored at -20\u0026deg;C prior to DNA isolation.\u003c/p\u003e\n\u003ch3\u003eMolecular characterization of leaf and airborne fungi\u003c/h3\u003e\n\u003cp\u003eFor both leaf and air samples, we added a sterile 6.35 mm chrome-steel bead (BioSpec Products, Inc., Bartlesville, Oklahoma, USA) to each microcentrifuge tube. The desiccated leaf material was pulverized dry while the air samples were pulverized wet, per kit instructions, using a TissueLyser II (Qiagen, Germantown, Maryland, USA) at 27 Hz for two minutes for each plate orientation (4 min total) per operating instructions.\u003c/p\u003e\u003cp\u003eDNA from the airborne samples was isolated using the DNeasy\u0026reg; PowerWater kit following the manufacturer\u0026rsquo;s protocol (Qiagen, Germantown MD, USA). We processed one DNA blank for every 47 samples during isolation. DNA from our leaf material was isolated using the Qiagen DNeasy\u0026reg; Plant Kit following the manufacturer\u0026rsquo;s protocol, with the following modifications. Because we had twice the maximum amount of dried sample in individual microcentrifuge tubes (~\u0026thinsp;200 mg) for the 96-well format, we doubled the amounts of AP1 and RNase and added them directly to each sample in its microcentrifuge tube in the heating step to ensure samples were at the correct concentration for adequate cell lysis and to improve RNA degradation. After vortexing, we transferred 400 \u0026micro;L of each lysate to its position in the 96-well plates, thereby adjusting the sample to its optimal concentration and volume and allowing us to proceed with the manufacturer\u0026rsquo;s protocol. We included the optional third ethanol wash step to further remove impurities from DNA. The DNA extracts were eluted from the column twice with 50 \u0026micro;L of Tris-EDTA and stored at -80\u0026deg;C.\u003c/p\u003e\u003cp\u003eFor leaves, we pooled 20 \u0026micro;L of DNA extract from each individual of a given species at each distance within each site into a single composite sample for that species at that distance at that site. Unfortunately, during this process six samples were inadvertently destroyed, one coffee sample and five non-Rubiaceae samples (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). Our negative control eluates were pooled as one sample prior to shipping. We shipped all eluates on dry ice to the University of Minnesota Genomic Center (UMGC, Minneapolis, MN, USA) for library preparation following the protocol in Sternhagen et al. (\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) and sequencing of the ITS2 region of the small ribosomal subunit using the fungal-specific primers ITS4 (White et al., \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e1990\u003c/span\u003e) and 5.8SR (Vilgalys \u0026amp; Hester, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e1990\u003c/span\u003e) and Illumina MiSeq\u0026trade; protocol in Gohl et al. (\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). UMGC conducted preliminary quality control and demultiplexing.\u003c/p\u003e\u003cp\u003eWe processed the raw FASTA sequences for the ITS2 region via the PIPITS 3.0 bioinformatics pipeline (Gweon et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) with a 97% similarity threshold, using the UNITE v8.3 database (Abarenkov et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). This fully automated pipeline prepares FASTA sequences, detects and removes chimeras and sequences\u0026thinsp;\u0026lt;\u0026thinsp;100 bp in length, assigns operational taxonomic units (OTUs), designates a taxonomic prediction and a confidence score for that prediction based on known sequences from the UNITE fungal database, and generates an OTU abundance matrix. We chose not to rarefy the OTU table to avoid discarding data (McMurdie \u0026amp; Holmes, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). Only OTUs with a kingdom confidence score of 1 (the highest possible score) were included in the final dataset. OTUs with taxonomic designations that had confidence scores\u0026thinsp;\u0026lt;\u0026thinsp;0.87 were considered unidentified. The highest read count for each OTU detected in the negative controls was subtracted from its read count, if any, in each of the biological samples.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eSampling effort across sites for image assessment of pathogen damage and molecular characterization of fungal communities in coffee, forest Rubiaceae, and forest non-Rubiaceae leaves. Numbers outside parentheses denote the number of leaves imaged and numbers inside parentheses denote the number of pooled molecular samples from these leaves. Fungal communities were characterized via ITS2 metabarcoding using the fungal-specific primers ITS4 (White et al., \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e1990\u003c/span\u003e) and 5.8SR (Vilgalys \u0026amp; Hester, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e1990\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"12\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c9\" colnum=\"9\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c10\" colnum=\"10\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c11\" colnum=\"11\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c12\" colnum=\"12\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/th\u003e\u003cth align=\"left\" colspan=\"8\" nameend=\"c10\" namest=\"c3\"\u003e\u003cp\u003eNumber of leaf samples by distance by site\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"2\" nameend=\"c12\" namest=\"c11\"\u003e\u0026nbsp;\u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSample type\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eDistance from coffee-forest border (m)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eAJ\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAN\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eDV\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eEP\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003eLZ\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e\u003cp\u003eRB\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c9\"\u003e\u003cp\u003eRN\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c10\"\u003e\u003cp\u003eVP\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c11\"\u003e\u003cp\u003eTotal\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eCoffee\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-100\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e3(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e-\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e28(7)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-50\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e8(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e37(8)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-25\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4(0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e33(7)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e-10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e4(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e33(8)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003eTotal\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e16(3)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cb\u003e16(4)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e20(4)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u003cb\u003e16(4)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e16(4)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e\u003cb\u003e15(4)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e\u003cb\u003e16(3)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e\u003cb\u003e16(4)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e131(30)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eForest Rubiaceae\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e12(4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e13(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e10(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e15(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e14(4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e12(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e8(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e16(5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e100(26)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e25\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e12(4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e13(4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e15(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e12(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e13(4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e20(5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e12(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e16(5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e113(31)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e50\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e17(4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e12(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e20(6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e12(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e11(4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e16(4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e13(4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e17(6)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e118(34)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e100\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e12(4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e12(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e19(5)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e13(4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e16(4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e21(4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e12(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e16(4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e121(31)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003eTotal\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e53(16)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cb\u003e50(13)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e64(16)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u003cb\u003e52(13)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e54(16)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e\u003cb\u003e69(16)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e\u003cb\u003e45(12)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e\u003cb\u003e65(20)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e452(122)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003e\u003cb\u003eForest non-Rubiaceae\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e10\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e8(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e11(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e9(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e12(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e12(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e12(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e8(0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e13(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e85(16)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e25\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e11(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e8(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e10(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e12(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e15(4)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e11(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e8(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e16(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e91(19)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e50\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e13(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e10(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e5(1)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e12(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e12(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e12(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e8(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e15(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e87(18)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e100\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e12(0)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e7(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e15(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e12(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e12(3)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e11(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e8(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e15(2)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e92(16)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u003cp\u003e\u003cb\u003eTotal\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u003cp\u003e\u003cb\u003e44(4)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u003cp\u003e\u003cb\u003e36(10)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cb\u003e39(7)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u003cp\u003e\u003cb\u003e48(11)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u003cp\u003e\u003cb\u003e51(12)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e\u003cp\u003e\u003cb\u003e46(11)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c9\"\u003e\u003cp\u003e\u003cb\u003e32(6)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c10\"\u003e\u003cp\u003e\u003cb\u003e59(8)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c11\"\u003e\u003cp\u003e\u003cb\u003e355(69)\u003c/b\u003e\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colspan=\"1\" nameend=\"c12\" namest=\"c12\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003eStatistical analyses\u003c/h2\u003e\u003cdiv id=\"Sec9\" class=\"Section3\"\u003e\u003ch2\u003eVisual assessment of foliar disease symptoms and fungal damage of seedlings\u003c/h2\u003e\u003cp\u003eWe conducted all analyses in R v4.2.3 (R Core Team, 2024). We used a two-step modeling approach to (1) determine whether leaf spot incidence was higher in coffee than in forest leaves, and (2) examine if leaf spot incidence changed with distance into the forest for forest Rubiaceae (native coffee relatives) versus forest non-Rubiaceae hosts. To test whether leaf spot incidence was greater in coffee than in forest leaves, we fitted a binomial generalized linear mixed model (GLMM) with presence/absence of leaf spot as the response, vegetation type (\u0026ldquo;coffee\u0026rdquo; vs. \u0026ldquo;forest\u0026rdquo;) as a fixed effect, and both site and plant species as random effects to account for site-to-site variation and variable sample sizes of given species. We used the Laplace approximation to ensure stable estimation despite sparse or uneven counts (Rue et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2009\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eTo determine whether close relatives of coffee (plants within Rubiaceae) exhibited a different pattern of edge-related leaf spot symptoms compared to more distantly related forest hosts, we classified each forest sample as belonging to either a Rubiaceae (close relatives of coffee) or a non-Rubiaceae host. We ran binomial GLMMs predicting leaf spot presence/absence as a function of the fixed effects physical distance from the coffee\u0026ndash;forest border and phylogenetic distance of the host species from coffee, plus their interaction, with site as a random effect for Rubiaceae and non-Rubiaceae forest plants.\u003c/p\u003e\u003cp\u003eWith respect to our seedling fungal damage surveys, we used a mixed linear regression model to explain change in fungal damage at the plot level as a function of physical distance from coffee field, phylogenetic distance from coffee, and their interaction term with forest site as a random effect to account for site variation.\u003c/p\u003e\u003c/div\u003e\u003c/div\u003e\n\u003ch3\u003eAirborne fungi\u003c/h3\u003e\n\u003cp\u003eTo evaluate whether total airborne fungal propagule load changed with distance from coffee into forest, we pooled monthly samples within sites to obtain an aggregated propagule load at each distance, avoiding potential pseudoreplication at each sampling location across months.We used a negative binomial mixed model on the counts along the transects with distance as a continuous fixed effect and site as a random effect. We used a beta-binomial mixed model with a logit link with the relative abundance of OTUs matched to the fungal order Pleosporales, which was the most abundant and diverse order detected on the coffee leaves (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e; Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) and includes many coffee pathogens, making it a strong candidate group for investigating coffee-to-forest spillover of fungi. These analyses were conducted using the \u0026ldquo;lme4\u0026rdquo; (v. 1.1\u0026ndash;31) and \u0026ldquo;glmmTMB\u0026rdquo; (v.1.1.12) packages (Bates et al., \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). We selected a beta-binomial approach because it can handle overdispersion and is commonly used to model compositional microbiome data (Martin et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003eFoliar fungi\u003c/h2\u003e\u003cp\u003eTo assess if host relatedness to coffee plays a role in the abundance of coffee-associated fungal taxa in the host phyllosphere, we calculated the phylogenetic distance of each of our forest plant species from coffee using methods described by Sternhagen (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Briefly, we acquired NCBI sequences spanning the 18S, ITS1, 5.8S, ITS2 and 26S regions of DNA for each of the plant species used in this study or the closest relatives with relevant sequence data (Table\u0026nbsp;2; Benson et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). All sequences we used contained at least a complete 5.8S segment. Phylogenetic distance from coffee for each understory plant species was extracted using the Tamura-Nei model with complete deletion via MEGA version 11.0.13 (Kumar et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). The phylogenetic distance of coffee to itself was set at zero.\u003c/p\u003e\u003cp\u003eTo evaluate if the relative abundance of Pleosporales declined from coffee into forest as with the air samplers, we fitted a beta-binomial mixed model, with the relative abundance of Pleosporales as the response variable, physical distance from the coffee-forest border as a fixed effect, and site as a random effect to partition variance due to sample location. To test whether the relative abundance of Pleosporales on forest leaves was influenced by proximity to the coffee field and the host plant's relatedness to coffee, we took a stratified approach and fitted the above model for forest Rubiaceae leaves and forest non-Rubiaceae separately. Phylogenetic distance of host plants from coffee was not a strong predictor of Pleosporales relative abundance in our Rubiaceae or non-Rubiaceae forest models and was removed from our final models. For final model selection, we compared all possible combinations of our variables of interest and selected the model with the lowest Akaike Information Criterion (AIC) value, following Burnham and Anderson (\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2004\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec12\" class=\"Section2\"\u003e\u003ch2\u003eInvestigating spillover of individual taxa\u003c/h2\u003e\u003cp\u003eFinally, to determine if individual fungal taxa from the air and leaves attenuated with distance into the forest, we first excluded taxa from the OTU matrix that appeared fewer than five times, as these would not yield a detectable signal across distances. We used generalized joint attribute modeling (GJAM; Clark et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2017\u003c/span\u003e) to test for fungal taxa that declined with distance into the forest interior using data from the forest air samplers (forest-facing at the border, and at 10, 25, 50, 100 m into the forest). We used physical distance from the coffee-forest border as a continuous fixed effect, sampling month as a categorical fixed effect, and site as a random effect. This modeling approach can accommodate multivariate datasets composed of various data types that are dominated by zeros in its probabilistic modeling approach (Taylor-Rodr\u0026iacute;guez et al., \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eFor forest leaf samples, the OTU matrix was too large to run in its entirety, even with the dimension reduction step built into this modeling approach. Consequently, we assessed the forest patterns of the 1,000 most abundant OTUs observed in coffee that also appeared on \u0026gt;\u0026thinsp;10 pooled leaf samples in the forest (~\u0026thinsp;56% of the total reads in the leaf samples), including all other OTUs as a binned \u0026ldquo;other\u0026rdquo; category in the model. We ran this full model with physical distance into the forest from the coffee-forest border and phylogenetic distance from coffee of the host plant as continuous fixed effects and site as a random effect. For both GJAM models we used three chains of 5,000 iterations and 500 burn-in iterations to achieve chain convergence. We matched the genera of OTUs to their ecological guild in order to determine the probable ecological role of these taxa and their potential as plant pathogens using the \u0026ldquo;Primary Lifestyle\u0026rdquo; designation of the FungalTraits database (P\u0026otilde;lme et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). For OTUs that could not be assigned to a genus, we manually assigned a guild designation when possible.\u003c/p\u003e\u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec14\" class=\"Section2\"\u003e\u003ch2\u003eLeaf spot on coffee and forest understory leaves and seedling fungal damage\u003c/h2\u003e\u003cp\u003eWe found the odds of leaf spot incidence were lower on forest understory plants than on coffee (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA). In forest Rubiaceae, the odds of leaf spot incidence declined with distance into the forest from the coffee\u0026ndash;forest border (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). However, for non-Rubiaceae plants in the forest understory, there was no effect of distance from the border on leaf spot incidence (\u003cem\u003eβ\u003c/em\u003e = \u0026minus;\u0026thinsp;0.06, \u003cem\u003ez =\u003c/em\u003e \u0026minus;\u0026thinsp;0.28, \u003cspan type=\"ItalicUnderline\" class=\"ItalicUnderline\" name=\"Emphasis\"\u003eSE\u003c/span\u003e\u0026thinsp;=\u0026thinsp;0.21, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.7790). There was no effect of phylogenetic distance from coffee on incidence of leaf spot in forest Rubiaceae or non-Rubiaceae (β = \u0026minus;\u0026thinsp;0.18, \u003cem\u003ez =\u003c/em\u003e \u0026minus;\u0026thinsp;0.76, SE\u0026thinsp;=\u0026thinsp;0.24, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.449; β\u0026thinsp;=\u0026thinsp;0.53, \u003cem\u003ez\u0026thinsp;=\u003c/em\u003e\u0026thinsp;1.66, SE\u0026thinsp;=\u0026thinsp;0.32, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0971; data not shown) and no interaction between distance and phylogenetic distance for either Rubiaceae or non-Rubiaceae (β\u0026thinsp;=\u0026thinsp;0.18, \u003cem\u003ez\u0026thinsp;=\u003c/em\u003e\u0026thinsp;0.725, SE\u0026thinsp;=\u0026thinsp;0.25. \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.4685; β = \u0026minus;\u0026thinsp;0.22, \u003cem\u003ez =\u003c/em\u003e \u0026minus;\u0026thinsp;0.98, SE\u0026thinsp;=\u0026thinsp;0.23, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.3284; data not shown).\u003c/p\u003e\u003cp\u003eAt the four sites where we surveyed seedling damage, there was a positive interaction between phylogenetic distance from coffee and physical distance from the coffee field in predicting fungal damage at the plot level (β\u0026thinsp;=\u0026thinsp;0.517\u0026thinsp;\u0026plusmn;\u0026thinsp;0.202, \u003cem\u003et\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.57, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.015; Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e), indicating that the negative effect of physical distance is strongest in hosts closely related to coffee and becomes progressively weaker in more distantly related hosts.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec15\" class=\"Section2\"\u003e\u003ch2\u003eAirborne fungi from coffee to forest\u003c/h2\u003e\u003cp\u003eFrom the 203 air samples we collected, the abundance table contained 2,465,390 reads and 6,838 OTUs. Total fungal read counts in the air increased with distance along the transect into the forest (β\u0026thinsp;=\u0026thinsp;0.005\u0026thinsp;\u0026plusmn;\u0026thinsp;0.002 SE, \u003cem\u003ez\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.26, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0236; Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). At the coffee-forest border, there were more read counts on average from air samplers facing the forest then those facing the coffee (β\u0026thinsp;=\u0026thinsp;2.62\u0026thinsp;\u0026plusmn;\u0026thinsp;1.27 SE, \u003cem\u003ez\u003c/em\u003e\u0026thinsp;=\u0026thinsp;2.07, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0389). In contrast, the relative abundance of Pleosporales declined significantly with increasing distance into the forest (β = \u0026minus;\u0026thinsp;0.012\u0026thinsp;\u0026plusmn;\u0026thinsp;0.003 SE, \u003cem\u003ez\u003c/em\u003e = \u0026minus;\u0026thinsp;3.85, \u003cem\u003eP\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0001) whether border-forest samplers were included or only forest samplers (Table S6).\u003c/p\u003e\u003cp\u003eAmong the fungal OTUs detected in the air, 33 declined in relative abundance from the coffee-forest border into the forest interior (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). These taxa included saprotrophs (13 OTUs), and potential plant pathogens/endophytes (9 OTUs), two entomopathogens, one epiphyte and one mycoparasite. Seven of the 33 could not be assigned to an ecological guild.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec16\" class=\"Section2\"\u003e\u003ch2\u003eFoliar fungi from the coffee-forest border into the forest interior\u003c/h2\u003e\u003cp\u003eAfter quality control steps, the OTU abundance table for foliar fungal communities contained 6,539,462 reads and 9,905 OTUs. The relative abundance of Pleosporales declined with physical distance into the forest on Rubiaceae but not non-Rubiaceae (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eB).\u003c/p\u003e\u003cp\u003eOf the 1,000 most abundant OTUs in coffee (hereafter \u0026ldquo;coffee-associated\u0026rdquo; taxa), 836 were observed on \u0026gt;\u0026thinsp;10 pooled sets of forest understory leaves, making up the majority (52.4%) of the total reads in the dataset. Of these 836 OTUs, 19 (\u0026lt;\u0026thinsp;3%) declined in relative abundance with increasing distance into the forest (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eA; Table S3). These included several taxa whose closest matches were potential plant pathogens (10 OTUs), which may also function as endophytes depending on host context, as well as closest matches to nematophagous (2 OTUs), saprotrophic (2 OTUs), a mycoparasitic (1 OTU), and lichenized (1 OTU) taxa.\u003c/p\u003e\u003cp\u003eAmong the 836 coffee-associated OTUs found on at least 10 molecular samples of forest understory plants, 93 (~\u0026thinsp;11%) exhibited greater relative abundance the more closely related the forest host species was to coffee (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eB). Among these taxa, nearly half of those assigned to an ecological guild were potential plant pathogens and endophytes (33 OTUs). The remaining taxa for which it was possible to assign a putative ecological guild included saprotrophs (25 OTUs), epiphytes and lichenized fungi (6 OTUs), mycoparasites (3 OTUs), and one nematophagous OTU (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eB; Table S4). An additional 33 OTUs could not be assigned to a guild. For 10 OTUs, there was an interaction of physical distance with phylogenetic distance, such that further from the edge, fungal read counts were less affected by the relatedness of the forest understory host plant to coffee (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eC).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eWe illustrate that coffee can act as a reservoir for foliar fungi, including plant pathogens, increasing disease pressure in adjacent forests for native plants closely related to the crop plant. Several lines of evidence support this: Visual surveys showed leaf spots were much more common on coffee leaves than on understory forest leaves. In forest Rubiaceae, the odds of leaf spot incidence declined with distance from the coffee\u0026ndash;forest border. Likewise, fungal damage on forest seedlings decreased sharply with distance from coffee fields in close coffee relatives, but this effect weakened in more distantly related hosts. Pleosporales, the dominant fungal order in coffee fields, were more abundant on forest Rubiaceae closer to the coffee\u0026ndash;forest border, mirroring their spatial pattern in airborne propagules. Lastly, among the 1,000 most abundant OTUs in coffee, roughly 10% were more abundant on plants the more closely related they were to coffee. The relative abundance of 19 of these OTUs, including many taxa with known potential as plant pathogens, declined with distance into the forest. This work demonstrates an underappreciated biotic edge effect of agricultural land adjacent to wildland habitats.\u003c/p\u003e\u003cp\u003ePrevious work on ecological spillover has focused on natural ecosystems as reservoirs of biodiversity that influence adjacent agricultural systems via pollinators (Garibaldi et al., \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2011\u003c/span\u003e), pests (Wisler \u0026amp; Norris, \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2005\u003c/span\u003e), or pathogens moving into crop system (Blitzer et al., \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2012\u003c/span\u003e). Our findings suggest this phenomenon can operate in reverse: intensive crop systems may act as reservoirs of crop-associated fungi, including those with pathogenic potential, that move into adjacent forest and increase disease pressure near habitat edges. This represents a biotic edge effect distinct from more commonly documented effects such as changes in light, humidity, temperature, canopy cover or plant density (Harper et al., \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; Willmer et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Over time, crop-driven spillover of pathogens could restructure adjacent forest plant communities by selectively disadvantaging plant taxa closely related to coffee. In this way, agricultural landscapes may not merely fragment habitats but actively restructure adjacent plant and fungal communities.\u003c/p\u003e\u003cp\u003eThe patterns we observed support our hypotheses that coffee-associated foliar fungi alter the foliar fungal communities of adjacent forest plants, particularly those closely related to coffee. For example, leaf spot, which was much frequent in coffee than forest, declined with distance into the forest away from the coffee-forest border only for forest Rubiaceae (Figs.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB, \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC). This pattern parallels the decline with distance into the forest of Pleosporales on Rubiaceae leaves (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eB), and in the air (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB). Indeed, the relative abundance of Pleosporales on forest Rubiaceae leaves was negatively associated with phylogenetic distance from coffee (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eC), suggesting that these fungi more readily colonize close relatives of the cultivated host. Together, these patterns support a host-specific spillover of foliar fungi from coffee into adjacent forest disproportionately affecting native Rubiaceae plants, facilitated by airborne coffee-associated fungal propagules\u003c/p\u003e\u003cp\u003eImportantly, spatial decline in disease was restricted to Rubiaceae; non-Rubiaceae hosts showed no such pattern. If microclimatic variation near the forest edge were the primary driver of elevated disease pressure, we would expect a uniform response across host lineages. Instead, the pattern was specific to Rubiaceae, suggesting a host-specific mechanism consistent with spillover from coffee, rather than changes in humidity, temperature, or light exposure acting as environmental filters (Laurance et al., 2002). Additionally, if microclimatic conditions alone were shaping leaf-associated fungal communities, we would expect a decoupling between airborne and leaf-surface communities (O\u0026rsquo;Malley, 2008; Wright et al., 2024). Instead, we observed parallel patterns for Pleosporales in air and leaves, suggesting Pleosporales propagule load in the coffee amplifies their abundance on compatible nearby forest hosts.\u003c/p\u003e\u003cp\u003eWe identified over 100 coffee-associated fungal taxa that appeared to exhibit host affinity for forest plants in the coffee family (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eB, \u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eC), a pattern consistent with known phylogenetic constraints on pathogen host ranges (Gilbert \u0026amp; Webb, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2007\u003c/span\u003e). Over 25% of these taxa had pathogenic potential, including members whose closest taxonomic match came from the genera \u003cem\u003eColletotrichum\u003c/em\u003e, \u003cem\u003eSetophoma\u003c/em\u003e, \u003cem\u003eCercospora\u003c/em\u003e, and the family Didymellaceae which contains pathogens (e.g. \u003cem\u003eColletotrichum kahawe, Setophoma terrestris, Cercospora coffeicola, and Phoma costaricensis)\u003c/em\u003e known to drive coffee leaf spot epidemics in Neotropical coffee fields (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eB; Table S4; Aveskamp et al., \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2010\u003c/span\u003e, Deb et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Aparecido et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; ). Intriguingly, in the subset of these taxa (~\u0026thinsp;10%, including 4 Didymellaceae taxa) in which the effect of phylogenetic distance depended on physical distance from the coffee field, the importance of phylogenetic distance diminished with increasing distance into the forest (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eC). This dissipating interaction suggests that while evolutionary relationships determine host susceptibility, realized colonization appears contingent on spatial proximity to the coffee field.\u003c/p\u003e\u003cp\u003eBroadly, these spillover dynamics could disrupt the density-dependent processes that structure forest understory communities. By increasing pathogen pressure near forest edges in proportion to the density of adjacent coffee plants, agricultural systems may impose a competing, externally driven density-dependent process capable of reshaping forest plant communities in fragmented landscapes. While the forest understory appears to harbor a broader consortium of airborne fungal propagules than coffee (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA), certain crop-associated groups, particularly Pleosporales, appear augmented in coffee and decline in abundance with distance into the forest (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB). This indicates that coffee acts not only as a source of high-density host-specific propagules but also selectively alters the composition of the fungal pool at forest edges.\u003c/p\u003e\u003cp\u003eJanzen\u0026ndash;Connell theory posits that host-specific enemies reduce survival near conspecific adults, limiting dominance by common species and allowing rarer species to persist, where plants at intermediate densities are most vulnerable to this effect (Connell et al., 1984; Bachelot et al., 2016). Yet the very mechanism that should promote the persistence of rare species closely related to the crop plant\u0026mdash;their low density\u0026mdash;may render these species most vulnerable at crop\u0026ndash;forest borders, where coffee-associated pathogens disproportionately infect close relatives of the crop. Conservation ecologists should explore how targeted management strategies, such as buffer vegetation composed of species distantly related to the crop (e.g., forage grasses between coffee fields and forest), could mitigate pathogen spillover into adjacent natural areas.\u003c/p\u003e\u003cp\u003eWhile the spatial distribution of fungal taxa is consistent with spillover from coffee, this work cannot directly trace the origin of fungal propagules found on or in forest plant tissues. Confirming that forest-colonizing foliar fungi originate in coffee fields requires directly tracking fungal movement via techniques like isotopic labeling or whole-genome sequencing, a critical next step for understanding the role of coffee fields as source populations for foliar fungi in surrounding landscapes. Another important follow-up would be to examine Rubiaceae seedling survival in forest fragments with or without adjacent coffee.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eHigh-density crop systems such as coffee fields can influence fungal communities in adjacent forests through ecological spillover, disproportionately affecting forest understory plants that are closely related to the crop plant. Our findings suggest that coffee fields serve as reservoirs for coffee-associated plant pathogens more likely to colonize nearby Rubiaceae plants in the forest understory. In the mosaic landscapes of the Neotropics. where agricultural land is in direct proximity to both forest fragments and protected forest areas, crop-mediated spillover of coffee pathogens into forest could alter forest plant community composition by amplifying disease pressure on native Rubiaceae. This presents an underappreciated biotic edge effect that can alter the plant community composition of apparently intact natural ecosystems. It is therefore crucial to promote larger, more contiguous conservation areas, particularly in the highly biodiverse tropics where agricultural encroachment is widespread and increasing.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003eAcknowledgements\u003c/p\u003e\n\u003cp\u003eWe are indebted to the farmers who allowed us access to their coffee farms and forests, without whom none of this work would have been possible. The staff at the Las Cruces Biological Station deserve much gratitude for creating a comfortable and productive place to work. The Organization for Tropical Studies (OTS) provided access to lab space and equipment as well as funding for food and lodging for GE from the Donald \u0026amp; Beverly Stone Fellowships (Fund 507/557). Enrique Castro Fonseca at OTS, and staff at the National Commission for the Management of Biodiversity (CONAGEBIO) in Costa Rica and USDA PPQ in Beltsville, Maryland, USA assisted with securing permits for research and collecting. We also thank Michael Atencio Picado and Erick Barrantes Fallas, who helped deploy and maintain the air samplers. This work was supported by the resources and staff at the University of Minnesota Genomics Center (https://genomics.umn.edu). The bioinformatic pipelines of this study were executed using the Center for Computationally Assisted Science and Technology (CCAST) at North Dakota State University, made possible in part by NSF MRI Award No. 2019077. This material is based upon work supported by the U.S. National Science Foundation Graduate Research Fellowship awarded to JAL (Grant No. 2137101), and NSF CAREER Award \u003cstrong\u003e2048131\u003c/strong\u003e and Award 2037121 to LAW.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eAbarenkov, K., Nilsson, R. H., Larsson, K.-H., Taylor, A. F. S., May, T. W., Fr\u0026oslash;slev, T. G., Pawlowska, J., Lindahl, B., P\u0026otilde;ldmaa, K., Truong, C., Vu, D., Hosoya, T., Niskanen, T., Piirmann, T., Ivanov, F., Zirk, A., Peterson, M., Cheeke, T. E., Ishigami, Y., \u0026hellip; K\u0026otilde;ljalg, U. (2024). The UNITE database for molecular identification and taxonomic communication of fungi and other eukaryotes: Sequences, taxa and classifications reconsidered. \u003cem\u003eNucleic Acids Research\u003c/em\u003e, \u003cem\u003e52\u003c/em\u003e(D1), D791\u0026ndash;D797. https://doi.org/10.1093/nar/gkad1039\u003c/li\u003e\n\u003cli\u003eAparecido, L. E. D. O., Loren\u0026ccedil;one, J. A., Loren\u0026ccedil;one, P. A., Torsoni, G. B., Lima, R. F. de, Chiquitto, A. G., Saqui, D., Oliveira J\u0026uacute;nior, G. G. de, \u0026amp; Rolim, G. de S. (2024). Climate change and \u003cem\u003ePhoma\u003c/em\u003e spp. leaf spot of Arabica coffee: A CMIP6 Modeling Approach. \u003cem\u003eRevista Brasileira de Meteorologia\u003c/em\u003e, \u003cem\u003e38\u003c/em\u003e, e38230041. https://doi.org/10.1590/0102-77863810041\u003c/li\u003e\n\u003cli\u003eApigo, A., \u0026amp; Oono, R. (2022). Plant abundance, but not plant evolutionary history, shapes patterns of host specificity in foliar fungal endophytes. \u003cem\u003eEcosphere\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(1), e03879. https://doi.org/10.1002/ecs2.3879\u003c/li\u003e\n\u003cli\u003eAveskamp, M. M., De Gruyter, J., Woudenberg, J. H. C., Verkley, G. J. M., \u0026amp; Crous, P. W. (2010). Highlights of the Didymellaceae: A polyphasic approach to characterise Phoma and related pleosporalean genera. \u003cem\u003eStudies in Mycology\u003c/em\u003e, \u003cem\u003e65\u003c/em\u003e, 1\u0026ndash;60. https://doi.org/10.3114/sim.2010.65.01\u003c/li\u003e\n\u003cli\u003eBates, D., Maechler, M., Bolker [aut, B., cre, Walker, S., Christensen, R. H. B., Singmann, H., Dai, B., Scheipl, F., Grothendieck, G., Green, P., Fox, J., Bauer, A., simulate.formula), P. N. K. (shared copyright on, Tanaka, E., \u0026amp; Jagan, M. (2025). \u003cem\u003elme4: Linear Mixed-Effects Models using \u0026ldquo;Eigen\u0026rdquo; and S4\u003c/em\u003e (Version 1.1-36) [Computer software]. https://cran.r-project.org/web/packages/lme4/index.html\u003c/li\u003e\n\u003cli\u003eBen\u0026iacute;tez-Malvido, J., L\u0026aacute;zaro, A., \u0026amp; Ferraz, I. D. K. (2018). Effect of distance to edge and edge interaction on seedling regeneration and biotic damage in tropical rainforest fragments: A long-term experiment. \u003cem\u003eJournal of Ecology\u003c/em\u003e, \u003cem\u003e106\u003c/em\u003e(6), 2204\u0026ndash;2217. https://doi.org/10.1111/1365-2745.13003\u003c/li\u003e\n\u003cli\u003eBenson, D. A., Cavanaugh, M., Clark, K., Karsch-Mizrachi, I., Ostell, J., Pruitt, K. D., \u0026amp; Sayers, E. W. (2018). GenBank. \u003cem\u003eNucleic Acids Research\u003c/em\u003e, \u003cem\u003e46\u003c/em\u003e(D1), D41\u0026ndash;D47. https://doi.org/10.1093/nar/gkx1094\u003c/li\u003e\n\u003cli\u003eBlitzer, E. J., Dormann, C. F., Holzschuh, A., Klein, A.-M., Rand, T. A., \u0026amp; Tscharntke, T. (2012). Spillover of functionally important organisms between managed and natural habitats. \u003cem\u003eAgriculture, Ecosystems \u0026amp; Environment\u003c/em\u003e, \u003cem\u003e146\u003c/em\u003e(1), 34\u0026ndash;43. https://doi.org/10.1016/j.agee.2011.09.005\u003c/li\u003e\n\u003cli\u003eBurnham, K. P., \u0026amp; Anderson, D. R. (2004). Multimodel Inference: Understanding AIC and BIC in model selection. \u003cem\u003eSociological Methods \u0026amp; Research\u003c/em\u003e, \u003cem\u003e33\u003c/em\u003e(2), 261\u0026ndash;304. https://doi.org/10.1177/0049124104268644\u003c/li\u003e\n\u003cli\u003eCadenasso, M. L., Pickett, S. T. A., Weathers, K. C., \u0026amp; Jones, C. G. (2003). A framework for a theory of ecological boundaries. \u003cem\u003eBioScience\u003c/em\u003e, \u003cem\u003e53\u003c/em\u003e(8), 750\u0026ndash;758. https://doi.org/10.1641/0006-3568(2003)053[0750:AFFATO]2.0.CO;2\u003c/li\u003e\n\u003cli\u003eCaudill, S. A., Brokaw, J. N., Doublet, D., \u0026amp; Rice, R. A. (2017). Forest and trees: Shade management, forest proximity and pollinator communities in southern Costa Rica coffee agriculture. \u003cem\u003eRenewable Agriculture and Food Systems\u003c/em\u003e, \u003cem\u003e32\u003c/em\u003e(5), 417\u0026ndash;427. https://doi.org/10.1017/S1742170516000351\u003c/li\u003e\n\u003cli\u003eClark, J. S., Nemergut, D., Seyednasrollah, B., Turner, P. J., \u0026amp; Zhang, S. (2017). Generalized joint attribute modeling for biodiversity analysis: Median-zero, multivariate, multifarious data. \u003cem\u003eEcological Monographs\u003c/em\u003e, \u003cem\u003e87\u003c/em\u003e(1), 34\u0026ndash;56. https://doi.org/10.1002/ecm.1241\u003c/li\u003e\n\u003cli\u003eDeb, D., Khan, A., \u0026amp; Dey, N. (2020). Phoma diseases: Epidemiology and control. \u003cem\u003ePlant Pathology\u003c/em\u003e, \u003cem\u003e69\u003c/em\u003e(7), 1203\u0026ndash;1217. https://doi.org/10.1111/ppa.13221\u003c/li\u003e\n\u003cli\u003eDelprete, P. G., \u0026amp; Jardim, J. G. (2012). Systematics, taxonomy and floristics of Brazilian Rubiaceae: An overview about the current status and future challenges. \u003cem\u003eRodrigu\u0026eacute;sia\u003c/em\u003e, \u003cem\u003e63\u003c/em\u003e, 101\u0026ndash;128. https://doi.org/10.1590/S2175-78602012000100009\u003c/li\u003e\n\u003cli\u003eGaribaldi, L. A., Steffan-Dewenter, I., Kremen, C., Morales, J. M., Bommarco, R., Cunningham, S. A., Carvalheiro, L. G., Chacoff, N. P., Dudenh\u0026ouml;ffer, J. H., Greenleaf, S. S., Holzschuh, A., Isaacs, R., Krewenka, K., Mandelik, Y., Mayfield, M. M., Morandin, L. A., Potts, S. G., Ricketts, T. H., Szentgy\u0026ouml;rgyi, H., \u0026hellip; Klein, A. M. (2011). Stability of pollination services decreases with isolation from natural areas despite honey bee visits. \u003cem\u003eEcology Letters\u003c/em\u003e, \u003cem\u003e14\u003c/em\u003e(10), 1062\u0026ndash;1072. https://doi.org/10.1111/j.1461-0248.2011.01669.x\u003c/li\u003e\n\u003cli\u003eGilbert, G. S., \u0026amp; Hubbell, S. P. (1996). Plant diseases and the conservation of tropical forests. \u003cem\u003eBioScience\u003c/em\u003e, \u003cem\u003e46\u003c/em\u003e(2), 98\u0026ndash;106. https://doi.org/10.2307/1312812\u003c/li\u003e\n\u003cli\u003eGilbert, G. S., \u0026amp; Webb, C. O. (2007). Phylogenetic signal in plant pathogen\u0026ndash;host range. \u003cem\u003eProceedings of the National Academy of Sciences\u003c/em\u003e, \u003cem\u003e104\u003c/em\u003e(12), 4979\u0026ndash;4983. https://doi.org/10.1073/pnas.0607968104\u003c/li\u003e\n\u003cli\u003eGohl, D. M., Vangay, P., Garbe, J., MacLean, A., Hauge, A., Becker, A., Gould, T. J., Clayton, J. B., Johnson, T. J., Hunter, R., Knights, D., \u0026amp; Beckman, K. B. (2016). Systematic improvement of amplicon marker gene methods for increased accuracy in microbiome studies. \u003cem\u003eNature Biotechnology\u003c/em\u003e, \u003cem\u003e34\u003c/em\u003e(9), 942\u0026ndash;949. https://doi.org/10.1038/nbt.3601\u003c/li\u003e\n\u003cli\u003eGolan, J. J., \u0026amp; Pringle, A. (2017). Long-distance dispersal of fungi. \u003cem\u003eMicrobiology Spectrum\u003c/em\u003e, \u003cem\u003e5\u003c/em\u003e(4). https://doi.org/10.1128/microbiolspec.FUNK-0047-2016\u003c/li\u003e\n\u003cli\u003eGoossens, D., \u0026amp; Offer, Z. Y. (2000). Wind tunnel and field calibration of six aeolian dust samplers. \u003cem\u003eAtmospheric Environment\u003c/em\u003e, \u003cem\u003e34\u003c/em\u003e(7), 1043\u0026ndash;1057. https://doi.org/10.1016/S1352-2310(99)00376-3\u003c/li\u003e\n\u003cli\u003eGweon, H. S., Oliver, A., Taylor, J., Booth, T., Gibbs, M., Read, D. S., Griffiths, R. I., \u0026amp; Schonrogge, K. (2015). PIPITS: An automated pipeline for analyses of fungal internal transcribed spacer sequences from the Illumina sequencing platform. \u003cem\u003eMethods in Ecology and Evolution\u003c/em\u003e, \u003cem\u003e6\u003c/em\u003e(8), 973\u0026ndash;980. https://doi.org/10.1111/2041-210X.12399\u003c/li\u003e\n\u003cli\u003eHarper, K. A., Macdonald, S. E., Burton, P. J., Chen, J., Brosofske, K. D., Saunders, S. C., Euskirchen, E. S., Roberts, D., Jaiteh, M. S., \u0026amp; Esseen, P.-A. (2005). Edge influence on forest structure and composition in fragmented landscapes. \u003cem\u003eConservation Biology\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e(3), 768\u0026ndash;782. https://doi.org/10.1111/j.1523-1739.2005.00045.x\u003c/li\u003e\n\u003cli\u003eHarvey, C. A., Pritts, A. A., Zwetsloot, M. J., Jansen, K., Pulleman, M. M., Armbrecht, I., Avelino, J., Barrera, J. F., Bunn, C., Garc\u0026iacute;a, J. H., Isaza, C., Munoz-Ucros, J., P\u0026eacute;rez-Alem\u0026aacute;n, C. J., Rahn, E., Robiglio, V., Somarriba, E., \u0026amp; Valencia, V. (2021). Transformation of coffee-growing landscapes across Latin America. A review. \u003cem\u003eAgronomy for Sustainable Development\u003c/em\u003e, \u003cem\u003e41\u003c/em\u003e(5), 62. https://doi.org/10.1007/s13593-021-00712-0\u003c/li\u003e\n\u003cli\u003eHoldridge, L. R., \u0026amp; Grenke, W. C. et al. (1971). Forest environments in tropical life zones: A pilot study. \u003cem\u003eForest Environments in Tropical Life Zones: A Pilot Study.\u003c/em\u003e https://www.cabdirect.org/cabdirect/abstract/19716605728\u003c/li\u003e\n\u003cli\u003eJohnson, B. L., \u0026amp; Haddad, N. M. (2011). Edge effects, not connectivity, determine the incidence and development of a foliar fungal plant disease. \u003cem\u003eEcology\u003c/em\u003e, \u003cem\u003e92\u003c/em\u003e(8), 1551\u0026ndash;1558. https://doi.org/10.1890/10-1072.1\u003c/li\u003e\n\u003cli\u003eKumar, S., Stecher, G., Li, M., Knyaz, C., \u0026amp; Tamura, K. (2018). MEGA X: Molecular Evolutionary genetics analysis across computing platforms. \u003cem\u003eMolecular Biology and Evolution\u003c/em\u003e, \u003cem\u003e35\u003c/em\u003e(6), 1547\u0026ndash;1549. https://doi.org/10.1093/molbev/msy096\u003c/li\u003e\n\u003cli\u003eLaurance, W. F., Nascimento, H. E. M., Laurance, S. G., Andrade, A., Ewers, R. M., Harms, K. E., Luiz\u0026atilde;o, R. C. C., \u0026amp; Ribeiro, J. E. (2007). Habitat fragmentation, variable edge effects, and the landscape-divergence hypothesis. \u003cem\u003ePLOS ONE\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e(10), e1017. https://doi.org/10.1371/journal.pone.0001017\u003c/li\u003e\n\u003cli\u003eMa, J., Li, J., Wu, W., \u0026amp; Liu, J. (2023). Global forest fragmentation change from 2000 to 2020. \u003cem\u003eNature Communications\u003c/em\u003e, \u003cem\u003e14\u003c/em\u003e(1), 3752. https://doi.org/10.1038/s41467-023-39221-x\u003c/li\u003e\n\u003cli\u003eMartin, B. D., Witten, D., \u0026amp; Willis, A. D. (2020). Modeling microbial abundances and dysbiosis with beta-binomial regression. \u003cem\u003eThe Annals of Applied Statistics\u003c/em\u003e, \u003cem\u003e14\u003c/em\u003e(1), 94\u0026ndash;115. https://doi.org/10.1214/19-AOAS1283\u003c/li\u003e\n\u003cli\u003eMcMurdie, P. J., \u0026amp; Holmes, S. (2014). Waste not, want not: Why rarefying microbiome data is inadmissible. \u003cem\u003ePLOS Computational Biology\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(4), e1003531. https://doi.org/10.1371/journal.pcbi.1003531\u003c/li\u003e\n\u003cli\u003ePajares-Murg\u0026oacute;, M., Garrido, J. L., Perea, A. J., L\u0026oacute;pez-Garc\u0026iacute;a, \u0026Aacute;., Bastida, J. M., \u0026amp; Alc\u0026aacute;ntara, J. M. (2024). Mutualistic and antagonistic phyllosphere fungi contribute to plant recruitment in natural communities. \u003cem\u003eJournal of Ecology\u003c/em\u003e, \u003cem\u003e113\u003c/em\u003e(6). https://doi.org/10.1111/1365-2745.14311\u003c/li\u003e\n\u003cli\u003ePendrill, F., Gardner, T. A., Meyfroidt, P., Persson, U. M., Adams, J., Azevedo, T., Bastos Lima, M. G., Baumann, M., Curtis, P. G., De Sy, V., Garrett, R., Godar, J., Goldman, E. D., Hansen, M. C., Heilmayr, R., Herold, M., Kuemmerle, T., Lathuilli\u0026egrave;re, M. J., Ribeiro, V., \u0026hellip; West, C. (2022). Disentangling the numbers behind agriculture-driven tropical deforestation. \u003cem\u003eScience\u003c/em\u003e, \u003cem\u003e377\u003c/em\u003e(6611), eabm9267. https://doi.org/10.1126/science.abm9267\u003c/li\u003e\n\u003cli\u003eP\u0026otilde;lme, S., Abarenkov, K., Henrik Nilsson, R., Lindahl, B. D., Clemmensen, K. E., Kauserud, H., Nguyen, N., Kj\u0026oslash;ller, R., Bates, S. T., Baldrian, P., Fr\u0026oslash;slev, T. G., Adojaan, K., Vizzini, A., Suija, A., Pfister, D., Baral, H.-O., J\u0026auml;rv, H., Madrid, H., Nord\u0026eacute;n, J., \u0026hellip; Tedersoo, L. (2020). FungalTraits: A user-friendly traits database of fungi and fungus-like stramenopiles. \u003cem\u003eFungal Diversity\u003c/em\u003e, \u003cem\u003e105\u003c/em\u003e(1), 1\u0026ndash;16. https://doi.org/10.1007/s13225-020-00466-2\u003c/li\u003e\n\u003cli\u003eRazafimandimbison, S. G., \u0026amp; Rydin, C. (2024). Phylogeny and classification of the coffee family (Rubiaceae, Gentianales): Overview and outlook. \u003cem\u003eTAXON\u003c/em\u003e, \u003cem\u003e73\u003c/em\u003e(3), 673\u0026ndash;717. https://doi.org/10.1002/tax.13167\u003c/li\u003e\n\u003cli\u003eReis Medeiros, H., Martello, F., Metzger, J. P., Harper, K. A., Mengual, X., Righi, C. A., \u0026amp; Ribeiro, M. C. (2022). Landscape composition regulates the spillover of beneficial insects between forest remnants and adjacent coffee plantations. \u003cem\u003ePerspectives in Ecology and Conservation\u003c/em\u003e, \u003cem\u003e20\u003c/em\u003e(2), 111\u0026ndash;116. https://doi.org/10.1016/j.pecon.2021.11.003\u003c/li\u003e\n\u003cli\u003eRice, R. A. (1999). A place unbecoming: The coffee farm of northern Latin America. \u003cem\u003eGeographical Review\u003c/em\u003e, \u003cem\u003e89\u003c/em\u003e(4), 554\u0026ndash;579. https://doi.org/10.1111/j.1931-0846.1999.tb00234.x\u003c/li\u003e\n\u003cli\u003eRue, H., Martino, S., \u0026amp; Chopin, N. (2009). Approximate Bayesian inference for latent Gaussian models by using integrated nested laplace approximations. \u003cem\u003eJournal of the Royal Statistical Society Series B: Statistical Methodology\u003c/em\u003e, \u003cem\u003e71\u003c/em\u003e(2), 319\u0026ndash;392. https://doi.org/10.1111/j.1467-9868.2008.00700.x\u003c/li\u003e\n\u003cli\u003eSpear, J. E., Grijalva, E. K., Michaels, J. S., \u0026amp; Parker, S. S. (2018). Ecological spillover dynamics of organisms from urban to natural landscapes. \u003cem\u003eJournal of Urban Ecology\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e(1), juy008. https://doi.org/10.1093/jue/juy008\u003c/li\u003e\n\u003cli\u003eSternhagen, E. C. (2019). \u003cem\u003eInfluences on Fungal Community Composition in Roots and Soil of Coffee and Other Rubiaceae in Costa Rica\u003c/em\u003e [North Dakota State University]. https://hdl.handle.net/10365/29307\u003c/li\u003e\n\u003cli\u003eSternhagen, E. C., Black, K. L., Hartmann, E. D. L., Shivega, W. G., Johnson, P. G., McGlynn, R. D., Schmaltz, L. C., Asheim Keller, R. J., Vink, S. N., \u0026amp; Aldrich-Wolfe, L. (2020). Contrasting patterns of functional diversity in Coffee Root Fungal Communities associated with organic and conventionally managed fields. \u003cem\u003eApplied and Environmental Microbiology\u003c/em\u003e, \u003cem\u003e86\u003c/em\u003e(11), e00052-20. https://doi.org/10.1128/AEM.00052-20\u003c/li\u003e\n\u003cli\u003eTaylor-Rodr\u0026iacute;guez, D., Kaufeld, K., Schliep, E. M., Clark, J. S., \u0026amp; Gelfand, A. E. (2017). Joint species distribution modeling: Dimension reduction using Dirichlet processes. \u003cem\u003eBayesian Analysis\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(4), 939\u0026ndash;967. https://doi.org/10.1214/16-BA1031\u003c/li\u003e\n\u003cli\u003eVilgalys, R., \u0026amp; Hester, M. (1990). Rapid genetic identification and mapping of enzymatically amplified ribosomal DNA from several Cryptococcus species. \u003cem\u003eJournal of Bacteriology\u003c/em\u003e, \u003cem\u003e172\u003c/em\u003e(8), 4238\u0026ndash;4246. https://doi.org/10.1128/jb.172.8.4238-4246.1990\u003c/li\u003e\n\u003cli\u003eWhite, T. J., Bruns, T., \u0026amp; Lee, S. (1990). Analysis of phylogenetic relationships by amplification and direct sequencing of ribosomal RNA genes. \u003cem\u003eA Guide to Methods and Applications\u003c/em\u003e.\u003c/li\u003e\n\u003cli\u003eWillmer, J. N. G., P\u0026uuml;ttker, T., \u0026amp; Prevedello, J. A. (2022). Global impacts of edge effects on species richness. \u003cem\u003eBiological Conservation\u003c/em\u003e, \u003cem\u003e272\u003c/em\u003e, 109654. https://doi.org/10.1016/j.biocon.2022.109654\u003c/li\u003e\n\u003cli\u003eWisler, G. C., \u0026amp; Norris, R. F. (2005). Interactions between weeds and cultivated plants as related to management of plant pathogens. \u003cem\u003eWeed Science\u003c/em\u003e, \u003cem\u003e53\u003c/em\u003e(6), 914\u0026ndash;917. https://doi.org/10.1614/WS-04-051R.1\u003c/li\u003e\n\u003cli\u003eZahawi, R. A., Duran, G., \u0026amp; Kormann, U. (2015). Sixty-seven years of land-use change in southern Costa Rica. \u003cem\u003ePLOS ONE\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e(11), e0143554. https://doi.org/10.1371/journal.pone.0143554\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Table 2","content":"\u003cp\u003eTable 2 is not available with this version.\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"nature-portfolio","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"","title":"Nature Portfolio","twitterHandle":"","acdcEnabled":false,"dfaEnabled":false,"editorialSystem":"ejp","reportingPortfolio":"","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-7781221/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7781221/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eAgricultural expansion creates mosaic landscapes where managed land borders undisturbed ecosystems, facilitating the ecological spillover of organisms between habitats. While fungal spillover from natural systems to crops is relatively well-studied, the reverse\u0026mdash;agricultural fungi colonizing adjacent forests\u0026mdash;remains underexplored. We investigated whether coffee fields act as reservoirs for foliar fungi that spill over into adjacent tropical forests, particularly affecting close relatives of coffee (Rubiaceae). Using transect-based field surveys, airborne spore sampling, and metabarcoding of fungal communities, we found that leaf spot incidence was much higher in coffee than in adjacent understory forest leaves and declined with distance into the forest for Rubiaceae plants. Pleosporales was the dominant fungal order in coffee leaves and declined in abundance with distance from the forest edge in both airborne samples and Rubiaceae leaves, suggesting propagule pressure from coffee fields led to changes in the phyllobiome of close coffee relatives in the adjacent forest. Among the 1,000 most abundant coffee-associated fungal taxa, 19 declined with distance into the forest, and 93 were more abundant on phylogenetically close relatives of coffee compared to non-Rubiaceae plants. Several of the closest matches for these taxa were potential plant pathogens. These results demonstrate that high-density crops can act as reservoirs for fungi capable of colonizing adjacent forest plants, disproportionately affecting plants closely related to the crop. This appears to disrupt natural plant pathogen dynamics in forests such that close relatives of the crop plant near the crop-forest border face increased disease pressure and may be at a competitive disadvantage as a result. These findings highlight an underappreciated edge effect of agriculture in fragmented landscapes.\u003c/p\u003e","manuscriptTitle":"Ecological spillover of coffee-associated foliar fungi from field to forest as a biotic edge effect in the Neotropics","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-28 23:52:25","doi":"10.21203/rs.3.rs-7781221/v1","editorialEvents":[],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"nature-communications","isNatureJournal":true,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"NCOMMS","sideBox":"Learn more about [Nature Communications](http://www.nature.com/ncomms/)","snPcode":"","submissionUrl":"https://mts-ncomms.nature.com/","title":"Nature Communications","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"Nature Communications","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"32595e2f-77cf-4484-873f-45a2fb301017","owner":[],"postedDate":"October 28th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[{"id":55960644,"name":"Earth and environmental sciences/Ecology/Forest ecology"},{"id":55960645,"name":"Earth and environmental sciences/Ecology/Agroecology"}],"tags":[],"updatedAt":"2025-10-28T23:52:25+00:00","versionOfRecord":[],"versionCreatedAt":"2025-10-28 23:52:25","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7781221","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7781221","identity":"rs-7781221","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}