Cost of resistance to herbivory in the annual plant Arabidopsis thaliana

This study examines the assumption that plant resistance to herbivory has fitness costs. To assess costs, a standard method was used for examining the significant negative genetic correlation between the resistance character and damage in the presence of herbivory and with fitness in the absence of herbivory. Seeds of five plants from four genotypes of Arabidopsis thaliana were sown under controlled conditions in a growth chamber. Half of the resulting two months-old rosettes were used for glucosinolate analysis and for herbivory assessment. The other half were transplanted into an enclosure in the natural habitat of this plant and their fitness (fruit number) were measured after harvesting the plants. Caterpillars from Spodoptera exigua were obtained from lab culture for herbivory assessment. Two second-instar caterpillars from S. exigua were placed on each of the rosettes. Larval weight of caterpillars was measured after 5 days. One hundred mg dry mass of leaves from 5 rosettes of each genotypes were used for HPLC analysis. There were genetic variations in types and quantities of glucosinolate between genotypes. The results from herbivory assessment showed that the larval weight of S. exigua fed on some genotypes was significantly lower than others, and therefore there was genetic variation in resistance to herbivore for A. thaliana genotypes. The statistical analysis showed that the larval weight of S. exigua was negatively correlated with the total glucosinolate concentration and with fruit number. Therefore, under the condition of this experiment, glucosinolates reduced damage by S. exigua and exhibited significant fitness costs.