Biodiversity analyses for risk assessment of genetically modified potato

An environmental risk assessment for the introduction of genetically modified crops includes assessing the consequences for biodiversity. In this study arthropod biodiversity was measured using pitfall traps in potato agro-ecosystems in Ireland and The Netherlands over two years. We tested the impact of site, year, potato genotype, and fungicide management regime on arthropod community composition. Three potato genotypes were compared: the cultivar Désirée, susceptible to the late blight pathogen Phytophthora infestans, a genetically modified cisgenic clone of Désirée resistant to P. infestans and the cultivar Sarpo Mira, also resistant to late blight. We aimed to test several ways to measure biodiversity in the context of risk assessment by using both univariate biodiversity indices and multivariate ordination methods, categorizing the pitfall trap catch by taxonomic or functional category. The Shannon-Wiener and Simpson biodiversity indices both showed strong differences between sites, years and potato genotypes, but showed no effects of the fungicide management regime. The effect of genotype was due to cultivar differences between Désirée and Sarpo Mira rather than between the GM-event (A15-31) and its isogenic comparator Désirée. Multivariate permutation analyses and RDA ordination confirmed these findings and also showed interactions between year, site and either genotype or treatment. The added value of the multivariate analysis was that it provided information on the specific arthropod groups or taxa that contributed to community structure. Multivariate analyses are recommended for use as a sensitive method to compare functionally important arthropod groups driving community structure within the framework of environmental risk assessments, or for the process of indicator species selection.