seed and plant characteristics in response to herbivore treatment

In the experiment Sinapis arvensis plants were exposed to one of four treatment: no herbivory (=ct), healthy Pieris brassicae caterpillars (=Pier), caterpillars parasitized by the gregarious Cotesia glomerata (=Cot), and caterpillars parasitized by the solitary Hyposoter eneninus (=Hyp). Plants were initially grown in the greenhouse but at the onset of flowering, they were transferred to tents placed in a garden to which pollinators were added. At this time point, the plants were exposed to the treatments. The experimental unit was a tray with 6 plants (=replicate), There were 4 replicates per treatment. When insect development was completed, we counted 1) the number of undamaged seed pods and 2) the number of damaged seed pods. Plants were then transferred to a garden in which the control plants were already present. When the seedpods had matured and dried, they were collected. Seed pods that had grown additionally were collected one month later (=new seed pods). Seeds were separated from their pods and weighed (=total seed mass in g). Per plant, cohorts of 20 seeds were weighed and their mean was calculated (=mean individual seed mass in microgram). We calculated the number of seeds per plants by dividing the weight of all seeds produced by a plant by its individual seeds mass (=number of seeds). After 2 years of storage in the dark, germination success of the seeds was determined by randomly choosing 50 seeds from each plant and allow them to germinate on wet filter paper. Germination success was calculated as the fraction of seeds that germinated out of 50. Reproductive potential was then calculated as the product of seed number and germination success. The data on the seed characteristics are in the ‘seed data per tray’ sheet, individual seed mass are in the sheet ‘seed mass of individual seeds) and the data on the seed pods are in ‘seed pods’ sheet