The semi-arid and arid zones of Australia – which cover 70% of the continent land mass – host a large diversity of habitats and ecological communities of high conservation importance. Among these, permanent and ephemeral wetlands are thinly dotted across the landscape, playing the key ecological roles of providing habitat, refugia and food resources to a wealth of aquatic and non-aquatic species. Arid-zone wetlands are of high spiritual, cultural and practical significance for aboriginal communities.
Wetlands of arid zones occur mostly as isolated pools scattered across the landscape. Occasionally, after (rare) intense rainfall or cyclonic events, stream channels refill and overflow, spilling out over floodplains, re-establishing connectivity among otherwise-isolated pools. These waterbodies host a plethora of invertebrate aquatic species characterized by great endemism and persisting through these isolation-connection cycles. Together, these species constitute unique communities of outstanding conservation importance.
The many factors that determine the distinctiveness of these invertebrate communities can be summarized into three main groups :
- Local ecological conditions such as water depth, temperature, salinity, oxygen content and presence of vegetation cover;
- Context and spatial configuration of the landscape of pools, for example, when topographic barriers limit hydrological connection and migration between waterbodies, and
- Dispersal abilities (species dispersal traits)– whether a species can disperse far enough across the inhospitable arid matrix by flying, swimming, hitch-hiking or catching the wind.
These three classes of factors are not mutually exclusive and their importance in explaining patterns of community composition might change over space and time.
In this study we investigated the factors driving the community composition of aquatic invertebrates in arid zones at different spatial scales. We used as case study the riverine wetlands of five drainage basins in the Pilbara region, in the NW of Australia (see pictures below). Using the riverine dendritic network as a basis, we generated a series of landscape models simulating various levels of landscape resistance to dispersal of invertebrate species. These accounted for potential connectivity between waterbodies under different river flow conditions, for topographic barriers and also for the willingness of a given species to depart the waterways during dispersal. For example, whereas dragonflies are able to fly long distances in any direction and even across the arid matrix, many mites depend on other species for transportation, and large crustaceans can only conquer new waterbodies when flow connection along the river network is re-established after rainfall events. We evaluated whether the dissimilarities in community composition between riverine waterbodies correlated with dissimilarity in environmental conditions, to the distance between pools, or both. We tested different distance measures between pools including the straight-line distance, convoluted paths via waterways, and the distance estimated from landscape resistance models that account for the difficulty in traversing terrain. We also tested for variation in spatial distribution patterns between groups of organisms with different modes of dispersal.
We found that the shape of the dendritic network in each basin, strongly constrained by the landscape context, seems to play an important role in determining community composition of riverine waterbodies. For example, where landscape barriers impede hydrological connectivity between streams and thus hinder dispersal of species, local environmental conditions correlated more strongly with community composition than any of the landscape distances and resistances tested. This was observed in two basins characterized by rough terrain and having a single main water channel with side branches that never get interconnected. On the other hand, in basins characterized by flat topography and extensive floodplains, community composition of aquatic invertebrates reflected with landscape resistance.
Animals with different modes of dispersal showed inconsistent patterns across the five drainage basins. We observed that in the basins with flatter topography, aerial dispersers tended to deviate from the main river channels, allowing occasional movement between nearby headwaters and resulting in similar communities in different river systems within the same basin. Surprisingly, strong flyers, despite their ability to fly across the arid matrix, are apparently guided along stream channels, perhaps due to greater moisture, food or likelihood of finding new habitat. This result highlights the importance of dry waterways for the dispersal of species and in enabling connectivity among waterbodies in arid environments. For obligate aquatic species, we did not find any clear pattern of spatial structure based on channel or landscape connectivity. This could relate to the fact that cyclonic events in the Pilbara refill stream channels, promoting high levels of dispersal and colonization of obligate aquatic species among sites within the same basin (especially in basins where streams can get connected laterally across floodplains), and therefore promote high levels of community evenness.
Overall, our study illustrates the complexity of the processes and factors that structure aquatic communities in extreme environments, such as the arid and semi-arid zones of Australia. It is important to understand and disentangle this complexity in order to optimize conservation investments (e.g. whether resources should be directed to preserve individual sites or networks of sites). This is of special relevance in areas like the Pilbara, where the ecological condition of wetlands is declining due to pastoralism, mining and other disturbances.
Morán-Ordóñez, A., Pavlova, A., Pinder, A., Sim, L., Sunnucks, P., Thompson, R.M. and Davis, J. (2015) Aquatic communities in arid landscapes: local conditions, dispersal-traits and landscape configuration determine local biodiversity. Diversity & Distributions, DOI: 10.1111/ddi.12342.