Abstract
In the past few decades, the rate of establishment of introduced plant species around the world has been accelerating. These plants can become invasive, dominating over native plants, resulting in lower discovery rates of native host plants by herbivores. For ovipositing insects, this can increase the chance of herbivore ‘mistake’ oviposition, where there is a preference to (or increased chance) lay eggs on a low fitness plant over higher fitness plants. This phenomenon is referred to as an ‘ecological trap’. Introduced or novel plants can cause ecological traps by being nutritiously unsuitable, having no physical defences and/or lacking unpalatable chemical compounds that the herbivore sequesters to provide protection from predators. This thesis aims to test if the Preference-Performance Hypothesis or an ecological trap applies to the endemic Vanessa gonerilla gonerilla (Kahukura, also known as the New Zealand Red Admiral), by studying their oviposition preference and larval performance of on native and introduced Urtica (nettles) of New Zealand.
Chapter 1 is a review of Lepidopteran oviposition choice, where I discuss studies corroborating and contradicting the Preference-Performance Hypothesis. According to the Preference-Performance Hypothesis, female insects should prefer to oviposit on a host plant that provides the highest fitness (larval performance) for her offspring. This chapter outlines the importance of studying the effects introduced plants can have on insect herbivores, and how ecological traps may occur when there is a mismatch in preference and performance. I discuss studies supporting the different hypotheses for Lepidoptera oviposition choice. Here I introduce the genus Urtica of New Zealand and the basic biology and cultural importance of V. g. gonerilla.
In Chapter 2 I use citizen science to determine where V. g. gonerilla are found in New Zealand, specifically in the wider Dunedin area, and which plants they are commonly seen on. The information from this study was then used when designing experiments for the remainder of the thesis. Vanessa gonerilla gonerilla were found most on a mix of native and introduced plant genera Veronica > Hedera > Urtica > Prunus = Cordyline = Achillea = Polypodiopsida. Basking and feeding on flowers were the most common behaviours observed. Vanessa gonerilla gonerilla were almost always seen feeding from the flowers of Veronica, suggesting this genus is an important nectar source for adults.
In Chapter 3, I investigated female V. g. gonerilla oviposition preference for native Urtica ferox, a native/introduced U. australis hybrid, and/or introduced U. urens using multichoice oviposition cages. When given the choice of Urtica species V. g. gonerilla chose to only lay eggs on U. ferox and U. australis hybrid, none laid eggs on introduced U. urens. Moreover, no butterflies chose to lay eggs on more than one Urtica species, in spite of individuals laying 10+ eggs each during the experiment. There was no difference in the number of eggs each individual chose to oviposit on U. ferox compared to U. australis. Therefore, our results support the hypothesis that V. g. gonerilla prefer to oviposit on native Urtica over introduced U. urens. Future studies could investigate coevolution and other mechanisms that may have led to this preference.
In Chapter 4 I tested the hypothesis that V. g. gonerilla larval development would vary based on host-Urtica diet in controlled laboratory conditions. Pupal weights were significantly higher for larvae reared on native U. ferox compared to those reared on native/introduced U. australis hybrid and introduced U. urens. There was no difference in pupal weights of those reared on U. australis hybrid compared to U. urens. Larvae raised on any of these host-Urtica species had the potential to reach adulthood; however, there was a higher probability of survival to adulthood when larvae were reared on U. australis.
In the final chapter, I discuss the results of chapters 2 through 4 in terms of how these V. g. gonerilla studies contribute to the larger picture. Because V. g. gonerilla oviposition preference ranking (U. ferox = U. australis > U. urens) qualitatively matches that of the larval development ranking (U. ferox > U. australis = U. urens), our findings support the Preference-Performance Hypothesis for this population of V. g. gonerilla. If they have the choice of native U. ferox and are not forced to use low fitness Urtica such as the introduced U. urens, these findings show no evidence of an ecological trap occurring within the Dunedin population of V. g. gonerilla. Through public outreach and promoting the use of citizen science platforms such as iNaturalist I hope to increase awareness of this beautiful taonga butterfly and the plants they rely on. The aim is to change the public perspective of V. g. gonerilla host plants, Urtica (stinging nettles), from a prickly dangerous weed, to recognising their value in the ecosystem and their own nuanced beauty.