The scaling of population growth of conifer invasions in New Zealand : A thesis submitted in partial fulfilment of the requirements for the Degree of Doctor of Philosophy at Lincoln University

Abstract

In ecology, we often want to answer questions about how populations change at broad spatial scales, but we frequently lack data at these scales. To fill this knowledge gap we typically extrapolate from small-scale observations, but common methods ignore spatial processes such as the neighbourhood effects of density and dispersal which have a large effect on plant populations. In this thesis I will use aerial imagery to develop new methods to measure scaling population growth. I will use this approach to show that broad scale processes slow population growth of conifer invasions in New Zealand. To do this, first I assessed the utility of using aerial imagery for detecting invasive conifers across different densities. Using data of trees measured in the field and detected from aerial imagery, I found that aerial imagery has a tendency to detect trees inconsistently across densities. However, for the mid-to-large-sized trees, this detection bias mostly disappeared and the imagery could reliably detect this size of trees. Due to this detection bias from density, I then examined the effects of detection bias on the estimates of scaling up population growth using simulations. I showed that even though there is a detection bias from aerial imagery, I am confident that this bias will not affect the estimates of population growth and scaling. Finally, using data collected from aerial imagery of eight conifer invasion sites across the South Island of New Zealand, I asked whether spatial processes have a large effect on the population growth of alien conifers. I expanded upon existing methodology to enable me to account for spatial processes in dynamic, rapidly changing systems. In doing so, I developed the method to factor in dispersal and the effects of density using a Bayesian spatial random effects model. I found that broad scale spatial processes slow population growth in conifer invasions. This study is the first of its kind to link data from aerial imagery to population dynamics models to test for the effects of spatial processes. Furthermore, a dataset of this scale, detail, and timespan has never before been collected for alien conifer invasions. By the end of this thesis, I will prove that spatial processes have a large effect on the population dynamics of invasive conifer trees. I show that when we scale up population growth from a small to broad spatial scale and we ignore the effects of space, we will vastly over-estimate the ability of a population to grow.