Sublethal effects of metal toxicity and the measure of plant fitness in ecotoxicological experiments

International audience

Английский. Anthropogenic pollution is a major driver of global environmental change. To be properly addressed, the study of the impact of pollutants must consider both lethal effects and sublethal effects on individual fitness. However, measuring fitness remains challenging. In plants, the total number of seeds produced, i.e. the seed set, is traditionally considered, but is not readily accessible. Instead, performance traits related to survival, e.g., vegetative biomass and reproductive success, can be measured, but their correlation with seed set has rarely been investigated. To develop accurate estimates of seed set, relationships among 15 vegetative and reproductive traits were analyzed. For this purpose, Noccaea caerulescens (Brassicaceae), a model plant to study local adaptation to metalcontaminated environments, was used. To investigate putative variation in trait relationships, sampling included several accessions cultivated in contrasting experimental conditions. To test their applicability, selected estimates were used in the first generation of a Laboratory Natural Selection (LNS) experiment exposing experimentally plants to zinc soil pollution. Principal component analyses revealed statistical independence between vegetative and reproductive traits. Traits showing the strongest positive correlation with seed set were the number of non-aborted silicles, and the product of this number and mean silicle length. They thus appeared the most appropriate to document sublethal or fitness effects of environmental contaminants in plant ecotoxicological studies. The relevance of both estimates was confirmed by using them to assess the fitness of parental plants of the first generation of an LNS experiment: the same families consistently displayed the highest or the lowest performance values in two independent experimental metal-exposed populations. Thus, both these fitness estimates could be used to determine the expected number of offspring and the composition of successive generations in further LNS experiments investigating the impact of multi-generational exposure of a plant species to environmental pollution.