Patterns and processes at multiple scales shape fish assemblage structure in tropical estuaries

A fundamental goal of ecology is to understand the mechanisms that regulate patterns of abundance, distribution, and richness of species' across landscapes. Achieving this will ultimately help to inform better designed ecological surveys, improve predictive capabilities, and enhance efficacy of management and conservation measures. Although coastal systems provide valuable nurseries for commercially and recreationally important nekton species around the world, conceptual frameworks to facilitate understandings of their faunal patterns are scant. However, it is clear that models of coastal ecosystem function are becoming increasingly sophisticated, progressing away from fine-scale, single-scale focuses to incorporate more of the processes that underpin patterns. In this thesis I use Australian tropical estuaries as model systems to further develop ideas of coastal ecosystem functioning, by demonstrating how a hierarchy of processes interact across a broad spectrum of scales to shape local faunal outcomes in a coastal system. Some information on broader-scale processes shaping faunal pattern in Australian tropical estuaries already exists. Starting at the broadest scale, biogeographic factors regulate species pools, setting the limits on which species can potentially utilise estuaries in a region. At a finer scale, within a bioregion the supply of recruits into individual estuary systems is systematically modified by the location of coastal spawning grounds relative to estuary mouths, the existing population size of self-recruiting estuary resident species, and connectivity to permanent freshwater recruit sources. To determine how recruits from the three different sources (coastal marine, within the estuary, and from permanent freshwater reaches) typically distribute at finer scales, within an estuary system, catch data were compared across three different reaches spanning the entire length of the river-estuary axis (lower estuary, transitional wetlands, freshwater reaches). Patterns of distribution were diverse within the assemblage, varying in a species- and life-history-specific manner, and emerging in 7 general 'modes of dispersal' along the estuary axis. Three of these modes describe varying levels of upstream dispersal by marine-spawned species, while an additional group of marine-spawned species were unexpectedly biased towards upstream reaches. The other 3 modes consisted of uniformly distributed estuary-residents, and two groups of freshwater species with varying levels of dispersal into the upper reaches of the estuary. The interfacing of these diverse 'modes of dispersal' means that habitats embedded in different reaches of the estuary will be subjected to very different species pools. Species pools in estuary reaches are not static however, but shift and morph seasonally in response to physico-chemical shifts and life-history cycles of estuary use. The nature and severity of seasonal shifts in faunal patterns were subsequently examined in the transitional zone of an estuary system, where the lower estuary and freshwater reaches interface. These transitional wetlands are the focus for extreme monsoon-driven physical shifts, and also subject to colonisation from all three recruitment sources. Fish were sampled on a monthly-bimonthly basis over 3 annual cycles, and trajectories of species' abundance and modal size-class revealed a diversity of temporal cycles that could be split into 4 modes based on varying responses to physical shifts and the relevance of transitional wetlands in lifehistories of species. This included: (1) classic nursery cycles of post-larval recruitment, growth, and emigration, (2) nursery cycles periodically interrupted by freshwater flows/floods, (3) recruitment delayed until after freshwater floods – presumably as the species initial recruit to ephemeral wetlands associated with floods, and (4) year-round wetland residence and selfrecruitment. These diverse and complex patterns suggest that assemblages will vary markedly relative to time of year sampled, as well as occurrence, timing and extent of monsoonal floods. Following floods, transitional wetlands fragment to a series of tidally connected pools, providing a tractable system to examine finer-scale processes shaping spatial structure of assemblages in a coastal wetland system. Twenty pools were sampled through two annual cycles, to assess the relative influence of local (i.e. environmental constraints) vs. regional drivers (i.e. dispersal processes) on assemblage structure. Faunal patterns suggested that assemblages were primarily structured according to the level of hydrological connectivity with the estuary channel, and secondarily by local environmental conditions in pools. The assemblage can be broken up into two components based on responses to connectivity: an estuary generalist component constrained by connectivity to better connected pools closer to the estuary channel, and a wetland specialist component that seemingly ascended gradients of elevation to access pools further from the channel, perhaps reflecting a drive to access a unique nursery habitat. Additionally, among lower elevation pools, where frequent connections facilitated redistribution, there was some evidence of species sorting relative to preferred conditions (e.g. depth, substrate type). These results illustrate how different patches of seemingly similar habitat may perform different functions for the assemblage due to their position in the landscape. To evaluate the extent to which spatial patterns in the wetland system may have been influenced by interactions with other faunal groups (prey), during the pre-wet season month of October in two consecutive years, benthic invertebrate and zooplankton assemblages were sampled in a subset of 13 pools, concurrently with fish surveys. Linkages between distribution of fish and invertebrate prey suggested that the major assemblage split across the wetland may have partially been a response to prey sources as well as a function of pool accessibility. Moreover, prey distributions explained some patterns among the better connected pools, exhibiting patterns consistent with hypotheses of bottom-up control. These results highlight the importance of biological interactions as a key component of the spatial ecology processes structuring fish assemblages in coastal wetlands. It is clear that local faunal patterns in Australian tropical estuaries are ultimately a function of all of these levels of process working in concert - processes characteristic of broader scales inevitably constrain faunal pattern at finer scales. Thus, in its simplest form this hierarchy of processes can be perceived as a succession of spatio-temporally variable filters imposed at different scales that sequentially refine the assemblage as levels are descended. Placing traditional study sites (or focal patches) within the framework of a broader ecosystem, recognising the interaction of processes across multiple scales in time and space, will therefore allow us to better account for observed patterns, and enhance the efficacy of ecological studies in these systems. The general principle of this hierarchical framework is also applicable to other coastal and estuarine systems in other parts of the world, although the exact nature of processes and their relative influence on faunal outcome will vary from place to place