SummaryThe potential impact of genetically modified (GM) crops on biodiversity is one of the main concerns in an environmental risk assessment (ERA). The likelihood of outcrossing and pollenmediated gene flow from GM crops and non-GM crops are explained by the same principles and depend primarily on the biology of the species. We conducted a national-scale study of the likelihood of outcrossing between 11 GM crops and vascular plants in Chile by use of a systematized database that included cultivated, introduced and native plant species in Chile. The database included geographical distributions and key biological and agronomical characteristics for 3505 introduced, 4993 native and 257 cultivated (of which 11 were native and 246 were introduced) plant species. Out of the considered GM crops (cotton, soya bean, maize, grape, wheat, rice, sugar beet, alfalfa, canola, tomato and potato), only potato and tomato presented native relatives (66 species total). Introduced relative species showed that three GM groups were formed having: a) up to one introduced relative (cotton and soya bean), b) up to two (rice, grape, maize and wheat) and c) from two to seven (sugar beet, alfalfa, canola, tomato and potato). In particular, GM crops presenting introduced noncultivated relative species were canola (1 relative species), alfalfa (up to 4), rice (1), tomato (up to 2) and potato (up to 2). The outcrossing potential between species [OP; scaled from 'very low' (1) to 'very high' (5)] was developed, showing medium OPs (3) for GM-native relative interactions when they occurred, low (2) for GMs and introduced noncultivated and high (4) for the grape-Vitis vinifera GM-introduced cultivated interaction. This analytical tool might be useful for future ERA for unconfined GM crop release in Chile.

A national‐scale study of outcrossing potential within Chilean vascular flora was conducted using an upgraded algorithm, which adds parameters such as pollinator agents, climate, and geographic conditions. Datasets were organized and linked in a Web platform (www.flujogenico.cl), in which the development of a total outcrossing potential (TOP) predictor was formulated. The TOP predictor is the engine in the Web platform, which models the effect of a type of agricultural practice on others (coexistence calculation mode) and on the environment (biodiversity calculation mode). The scale for TOP results uses quintiles in order to define outcrossing potential between species as “very low,” “low,” “medium,” “high,” or “very high.” In a coexistence analysis considering 256 species (207 genera), the 10 highest TOP values were for genera Citrus, Prunus, Trifolium, Brassica, Allium, Eucalyptus, Cucurbita, Solanum, Lollium, and Lotus. The highest TOP for species in this analysis fell at “high” potential, 4.9% of the determined values. In biodiversity mode, seven out of 256 cultivated species (2.7%) were native, and 249 (97.3%) corresponded to introduced species. The highest TOP was obtained in the genera Senecio, Calceolaria, Viola, Solanum, Poa, Alstroemeria, Valeriana, Vicia, Atriplex, and Campanula, showing “high” potential in 4.9% of the values. On the other hand, 137 genetically modified species, including the commercial and pre‐commercial developments, were included and represented 100 genera. Among these, 22 genera had relatives (i.e., members of the same genus) in the native/introduced group. The genera with the highest number of native/introduced relatives ranged from one (Ipomea, Limonium, Carica, Potentilla, Lotus, Castanea, and Daucus) to 66 species (Solanum). The highest TOP was obtained when the same species were coincident in both groups, such as for Carica chilensis, Prosopis tamarugo, and Solanum tuberosum. Results are discussed from the perspective of assessing the possible impact of cultivated species on Chilean flora biodiversity. The TOP predictor (http://epc.agroinformatica.cl/) is useful in the context of environmental risk assessment.