The enemy release hypothesis is often cited as a potential explanation for the success of introduced plants; yet empirical evidence for enemy release is mixed. We aimed to quantify changes in herbivory and defence in introduced plants while controlling for three factors that might have confounded past studies: using a wide native range for comparison with the introduced range, measuring defence traits without determining whether they affect herbivore preferences, and not considering the effect of time since introduction. The first hypothesis we tested was that introduced plants will have evolved lower levels of plant defence compared to their source population. We grew South African (source) and Australian (introduced) beach daisies (Arctotheca populifolia) in a common-environment glasshouse experiment and measured seven defence traits. Introduced plants had more ash, alkaloids and leaf hairs than source plants, but were also less tough, with a lower C:N ratio and less phenolics. Overall, we found no difference in defence between source and introduced plants. To determine whether the feeding habits of herbivores align with changes in defence traits, we conducted preference feeding trials using five different herbivore species. Herbivores showed no overall preference for leaves from either group. The second hypothesis we tested was that herbivory on introduced plant species will increase through time after introduction to a new range. We recorded leaf damage on herbarium specimens of seven species introduced to Eastern Australia, and three native control species. We found no change in the overall level of herbivory experienced by introduced plants since arriving in Australia. Conclusion: In the field of invasion ecology we need to rethink the paradigm that species introduced to a new range undergo simple decreases in defences against herbivores. Instead, plants are likely to employ a range of defence traits that evolve in both coordinated and opposing ways in response to a plethora of different biotic and abiotic selective pressures.

Data for the six leaf traits (leaf toughness, ash, alkaloids, phenolics, carbon to nitrogen ratio and leaf hairs) were collected using standard protocols from South African source and Australian introduced Arcthotheca populifolia grown in a glasshouse at UNSW. Further methods are available in the published manuscript.

Data for herbivore preference were collected by offering four herbivore species [green garden looper caterpillar (Chrysodeixis sp.), garden snail (Cornu aspersum), grey field slug (Deroceras reticulatum) and lily caterpillar (Spodoptera picta)] a choice of a South African source leaf or an Australian introduced leaf and then measuring the leaf damage. Data were also gathered by counting the number of red spider mites (Tetranychus urticae) on the two groups of plants. Further methods are available in the published manuscript.

Data for herbivory through time were obtained by recording the percent of leaf area lost in herbarium specimens of seven species introduced to eastern Australia. Further methods are available in the published manuscript.

There are data for four Australian populations and one South African population. This is because we used genetic methods to identify the South Afrcian source population for comaprison with the introduced Australian plants.

More details are available in: Brandenburger, C.R., Sherwin, W.B., Creer, S.M., Buitenwerf, R., Poore, A.G., Frankham, R., Finnerty, P.B. & Moles, A.T. (2019) Rapid reshaping: the evolution of morphological changes in an introduced beach daisy. Proceedings of the Royal Society B, 286, 20181713. (https://doi.org/10.1098/rspb.2018.1713)