Plant functional traits associated with shoot flammability: A thesis submitted in partial fulfilment of the requirements for the degree of Doctor of Philosophy in Ecology at Lincoln University

Abstract

Fire is a common natural disturbance in many parts of the world, which changes species composition and ecosystem processes, shaping many of the world’s biomes. Fire has been an important disturbance for hundreds of millions of years, since the origin of terrestrial plants. Several factors interact to determine fire behaviour, including the prevailing weather conditions, the topography of the landscape, and the fuels available for the fire to burn. The main fuels available to initiate and sustain wildfires are plants. Hence, to understand wildfire behaviour, it is essential to know how easily plants ignite and how well they burn; that is, we must be able to measure plant flammability. Flammability is a complex plant trait that is not easy to measure and which differs across different plant parts or components. Flammability studies conducted on small plant components in the laboratory may not reflect how a plant will burn in nature. Furthermore, flammability research lacks a standard way of quantifying plant flammability in the laboratory. In this thesis, I seek to help address these shortcomings and gain a better understanding of plant flammability, first by comparing the flammability measurements at two different plant components (leaves and shoots) in the laboratory. Next, I investigated the influence of leaf morphological, chemical, and fuel architectural traits on flammability. Finally, I examined the role of species mixtures in causing variation in flammability, by burning fuel mixtures consisting of species of varying levels of flammability. Comparing the flammability of fuels from different plant components (leaf- and shoot-level) (Chapter 2), I showed that flammability measurements between these fuels were uncorrelated and provided evidence that shoot flammability is likely to be better than leaf flammability at estimating the flammability of plants in the field, at least for canopy fuels. When investigating the relationship between leaf traits and shoot flammability using leaf-level morphological and chemical traits of 43 species collated from trait databases (Chapter 2), I demonstrated that leaf traits such as leaf dry matter content (LDMC), leaf thickness, leaf phenolics, and leaf lignin were correlated with shoot flammability, and thus have potential as useful and easily-measured surrogates for flammability. To further investigate trait flammability relationships, I measured leaf and architectural traits of 65 indigenous and exotic New Zealand species, along with shoot flammability, on the same individuals of a species (Chapter 3). I provided further evidence that LDMC and leaf thickness were strongly correlated with shoot flammability, and that branching pattern (number of ramifications and sub-branches) was the most important architectural trait influencing shoot-level flammability. Other architectural traits, such as foliage and twig fraction mass and fuel bulk density, were also shown to be correlated to shoot flammability. Given that fires often burn through vegetation that contains plant species of varying flammability, I investigated how the flammability of fuel mixtures was affected by the flammability of constituent species. Using shoot samples from two high flammability and two low flammability species (Chapter 4), I showed that the flammability of species mixtures was non-additive (i.e. disproportionately influenced by the flammability of the constituent species), and that the low flammability species significantly reduced flammability variables, such as burning time and total heat release of the species mixture. I have demonstrated that shoot-level flammability measurements represent a more suitable laboratory-based means of quantifying canopy flammability than the more widely-used approach of measuring leaf-level flammability. Furthermore, I have quantified the effects of leaf and architectural traits on shoot flammability and identified key traits, such as LDMC, which can be used as surrogates for plant flammability. Finally, I have identified the role of plant species of varying flammability in changing the fire behaviour of species mixtures, demonstrating the mechanisms by which low flammability plants can be used to reduce fire impacts. These findings not only contribute to a greater understanding of how plants burn, they should prove useful for model-based approaches to predicting changes to fire regimes, and provide crucial information to fire managers seeking to mitigate fire damage in an increasingly fire-prone world.