Coffee plantation fungi spill into Costa Rican forest

A plantation boundary does not stop spores. A team working in Costa Rica tested whether fungi associated with coffee cultivation also appear on plants in the neighbouring forest. The researchers combined observations of leaf spotting, sampling from plants and analysis of fungal DNA along transects running from the coffee fields into the forest. The result indicates a biological edge effect: the influence of cultivation was detectable beyond its physical boundary.

The clearest signal concerned plants in the family Rubiaceae, to which coffee belongs. The frequency of leaf spotting on forest Rubiaceae decreased with distance from the plantation. Some groups detected by molecular methods showed a similar pattern, including fungi in the order Pleosporales. The authors also identified nineteen taxa associated with coffee whose presence in the forest declined with increasing distance from the field. That is consistent with propagules being carried through the air and colonising closely related hosts.

This does not mean that every fungus detected caused disease, or that the plantation "infected the whole forest". A DNA sequence confirms the presence of an organism or its material, but does not by itself establish whether it is an active pathogen. Spots on leaves can also have many causes. The strength of the study lies in the convergence of several independent signals and their spatial arrangement; the limitation is that one landscape and one cropping system do not represent all tropical plantations.

The authors also modelled the frequency of taxa on different forest plants. For ninety-three taxa the predicted abundance was higher on wild Rubiaceae than on plants of other families. Host relatedness may therefore ease the passage from crop to forest. That has practical significance, because many plantations lie within a mosaic of habitats, and a strip of forest is often assessed mainly as a buffer for animals, water or soil. Microorganisms show the less visible side of that boundary.

The conclusion is not "remove coffee from the neighbourhood of forest", but that landscape design should also take account of the flow of pathogens and other fungi. The width of a buffer zone, the composition of plants at the edge, the crop varieties and the way a plantation is managed may change the risk, but each has to be studied separately. For a consumer it is a reminder that the impact of an agricultural product does not end at the edge of a field. For ecologists, that a boundary visible on a map need not be a biological boundary. The next step should be repeatable studies across many farms, seasons and shade systems, combined with tests showing which of the detected fungi actually infect forest plants.

Evidence status — 29 August 2026

The peer-reviewed study was published on 24 July 2026. In one Costa Rican landscape it combined leaf-spot observations, fungal DNA identification and transects running from coffee into forest. Nineteen coffee-associated taxa declined in occurrence with distance from the plantation, with a particularly clear signal on wild Rubiaceae. That spatial pattern is consistent with fungal spillover beyond the crop, but a DNA read does not prove that every detected taxon was an active pathogen. Replication across farms and seasons is still needed. The result already matters because it shows that a crop’s biological edge need not coincide with the boundary drawn on a map.