Metacommunity structure and connectivity in dendritic ecological networks

Abstract

Many freshwater fish communities are declining globally, one of the primary causes of which is changes to connectivity regimes. Connectivity strongly influences the ability of organisms to move around the landscape. In riverine ecosystems, network topology and spatio-temporal environmental factors, such as flow variability and the presence of barriers, determine connectivity. Furthermore, connectivity loss and processes such as climate change do not influence biodiversity independently but will likely affect populations and communities synergistically. Consequently, my thesis was motivated by the need to improve our understanding of freshwater fish population and community dynamics in the face of changing connectivity and environmental regimes. In this thesis I aim to address some of the knowledge gaps in these areas by applying the theoretical principles of metacommunity ecology in conjunction with graph-theoretic methods. First, I analysed the utility of graph-theoretic metrics for quantitatively describing simulated and real dendritic ecological networks, finding that node scale metrics do not respond in a predictable way to algorithm input parameters and that simulated networks are topologically different to real rivers regarding node metrics. I then used a discrete-time logistic growth metacommunity model to analyse the role of spatial and temporal functional connectivity in determining patterns of local species richness in freshwater fish metacommunities. The results of this modelling suggest that: i) in dendritic ecological networks environmental spatial structure can determine how communities respond to disturbance; ii) the effect of spatial loss of connectivity on local species richness is determined by network topology and where richness is being measured; and iii) increasing temporal autocorrelation in connectivity results in increasing temporal autocorrelation in patch occupancy. Finally, I examined how intercatchment connectivity can affect extinction risk via source-sink dynamics, using the upokororo (Prototroctes oxyrhynchus) as a case-study. I gathered historical data on the species to recreate a distribution map for the species and to predict a likely extinction horizon. I then modelled the species’ population dynamics to show that by accounting for amphidromous dispersal it is possible to explain how the species may have gone extinct under relatively light harvesting pressure. Overall, I conclude network topology and functional connectivity affect patterns of persistence and species richness in their own right. Further, the effects of other drivers of population and community dynamics, such as environmental variability can depend on underlying connectivity conditions.