Defining plant functional groups to guide rare plant management

Abstract

To effectively manage plant populations for conservation, there is a need to provide reliable information on the conditions required for maintaining viable populations. This is particularly true for the management of populations of rare plant taxa. Western Australia contains over 45% of Australia’s gazetted rare or threatened flora, 80% of which are found within the highly fragmented southwest region. Resources do not exist to undertake comprehensive studies on the population dynamics and demographics for every rare plant of this diverse region. Here, we describe a method of classifying rare plant taxa into functional groups as a basis for guiding rare flora conservation and management. Data on four floral and two life-history traits were collected for each of the 351 declared rare flora taxa of Western Australia. A hierarchical, agglomerative clustering method was applied to the resulting taxa by traits matrix to extract emergent groupings of plant taxa. The resulting polythetic groups were analysed to determine the variation in traits, including response to disturbance and recorded flower visitors, and how these may affect population persistence in a fragmented landscape. Multivariate methods were used to define emergent groups based on a combination of floral structure and life-history traits of the declared rare flora of Western Australia. Seven emergent functional groups were identified and were largely differentiated by flower shape and life form. These seven functional groupings varied significantly in their response to disturbance. By deriving these functional groups, we plan to develop models for each group on how rates of pollination, seed production and seed fitness are affected by population size and landscape context. The rationale would be to use these profiles to determine whether there are thresholds in population size or position in the landscape at which reproductive rates severely decline. General management guidelines could then be developed for each functional group.