Search

Conservation planning processes and wetland management require spatial estimations of aquatic habitats to support the maintenance of aquatic biodiversity. However, physical access to several wetlands and freshwater habitats can be restricted due to difficult topography and technological limitations associated with ground-based observations. In addition to these constraints, the distribution of some aquatic primary producers in freshwater habitats and floodplains that are difficult to reach further complicates large-scale assessment of aquatic habitats using traditional field-based techniques. The main objective of this study is to predict the spatial distribution of hot spots of primary producers (aquatic plant biomass) and assess large-scale inundation patterns in a large floodplain wetland (Flinders catchment) in the wet-dry tropics of Australia. To this end, remote sensing biophysical indicators (vegetation and inundation) were integrated with flood water depth in a classification tree model. Results indicate that in terms of floodplain total inundation, the Flinders wetland hydrology is rather restricted immediately after the summer wet season, the period when most primary production happens. While this can be attributed to the fact that much of the observed annual variability (93%) in rainfall and surface runoff (95%) occur during the wet season, post flood recession patterns are indicators that underpin the somewhat limited alimentation of the Flinders floodplain during this period. As observed in this study, post flood inundation extents in the summers of 2009 and 2019 declined by approximately 89% and 87% within fourteen and ten days, respectively. Despite having a significantly higher magnitude than the 2009 summer flood event, the 2019 extreme 'big wet' period did not translate to higher floodplain productivity (aquatic plant biomass and surface water distribution) immediately after summer as was the case in 2009. Furthermore, the predicted extents of aquatic plant biomass and total floodplain inundation in the downstream Flinders show substantial temporal variation and suggest the floodplain wetland hydrology is largely driven by inter-annual changes in annual rainfall. The extents of these hot spots of biomass accumulation were found to be considerably associated with total floodplain inundation extent (r = 0.94), rainfall (r = 0.81), and discharge (r = 0.68). Furthermore, advance statistical analyses show that downstream discharge and rainfall over the Flinders are significantly correlated (r = 0.72). In addition to this, observed amplitudes of discharge and rainfall in extreme wet and dry years coincided with floodplain inundation patterns and distribution of hot spots of primary producers. While these relationships emphasize the importance of flow in the nourishment of the downstream catchment, they further corroborate the composite influence of local rainfall and discharge on total floodplain inundation and hot spots of primary producers. [ABSTRACT FROM AUTHOR]

• Introduced a cloud-based framework to assess changes in floodplain productivity. • Upstream flows drive floodplain inundation and aquatic plant biomass accumulation. • Local rainfall has limited impact on the productivity of downstream catchment. • Climate variability is a key constraint to changes in floodplain productivity. • Flows are primary drivers of floodplain inundation in the Mitchell catchment. Rivers in tropical systems across the world are well known for their strong connections with floodplain wetlands. However, increased water needs and changing climate could drive water-management policy that leads to major changes in flow or surface water availability. These changes could have deleterious effects on ecological functions and ecosystem services provided by these freshwater ecosystems (e.g., groundwater recharge, food sources for fish, etc.). In this study, we introduce a cloud-based monitoring framework using multi-resolution satellite observations (2007 - 2020), historical surface inundation (1984 - 2015), and gauge data to better understand these connections in a large river catchment in tropical Queensland (Mitchell River catchment). A key driver and indicator of floodplain productivity, surface water inundation and aquatic biomass distribution respectively, are investigated to provide insight that underpins the prioritization of future surface water developments in this region. Results show that the characteristics of floodplain inundation; loss, persistence, gain and frequency, vary in time and space, and are driven by climate variability. The largest gain (7200 Ha) and persistence (8437 Ha) in total surface inundation were observed in the wet seasons during the periods of 2007 - 2019 , respectively. In the downstream catchment, permanent water surface area was approximately 4.5% (of the total downstream region) compared to the upstream region (2.7%). There is a higher proportion of new permanent water features (16.8%) and new seasonal (30%) surface water classes in the upstream are in contrast to the downstream section (0.7% and 9.6%, respectively). Using multivariate statistics, floodplain productivity was linked to inter-annual changes in flow and rainfall. A partial least squares regression (PLSR) model shows upstream flows drive changes in total floodplain inundation (observed vs predicted, r = 0.95 , p = 0.000). Flows and local rainfall accounted for approximately 60% (r = 0.77 , p = 0.005) and 25% (r = 0.50 , p = 0.120) of the variability in total floodplain inundation, respectively. Different PLSR modelling scenarios confirmed that upstream flows are primary drivers of floodplain inundation and substantial lost in total surface inundation could happen if flows are altered. Just as local rainfall has limited impact on total floodplain inundation (r = 0.92 , p = 0.000 ; for model excluding local rainfall), aquatic plant biomass accumulations are also strongly associated with upstream flows (e.g., r = 0.82). Given the dependence of downstream water level (r = 0.89) and floodplain inundation on upstream flows, sustainable management of river flows in the Mitchell River is fundamentally important to floodplain productivity. [ABSTRACT FROM AUTHOR]

• Floodplain productivity and inundation assessed using remote sensing indicators. • River discharge is a stronger indicator of hot spots of biomass accumulation. • Floodplain inundation and primary production are driven by flow and local rainfall. • Upstream catchment contributes to the connectivity and productivity of floodplain. • Fluctuations in climate will be key constraints on freshwater habitats. The ecological communities supported by freshwater habitats and wetlands that persist from floodplain inundation generate numerous cultural, recreational and economic values via commercial fisheries and other human uses of these habitats. However, the alteration of flow connectivity, degradation, and disruption of physical processes that sustain different levels of organisms (e.g., primary producers) are threats that affect aquatic biodiversity and the productivity of these habitats. Therefore, the large scale assessment of freshwater habitats for proactive water resources planning and the development of climate change mitigation and resource management strategies is essential. However, such monitoring and assessment is complicated by the inaccessibility of many large wetland systems during times of inundation, making in situ sampling unfeasible at a time when high levels of aquatic primary producers are generating food and energy sources for higher order consumers. To understand the physical dynamics of aquatic primary producers and freshwater habitats (1991–2019) during such times in a large floodplain river in northern Australia (Gilbert catchment), this study integrated Landsat-derived modified normalised difference water index and normalised difference vegetation index in a classification tree model. Thereafter, key hydrological drivers (rainfall and river discharge) of floodplain productivity and connectivity were assessed using a range of multivariate techniques. Results show that the floodplain had a high aquatic plant biomass during the summer wet season based on the correlation between aquatic biomass accumulation (hot spots) and inundation (r = 0.83 @ phase lag < 1 month) during such period. River discharge (r = 0.68 @ lag = 1 month) at downstream Gilbert catchment appear to be a relatively stronger indicator of hot spots of floodplain biomass accumulation as opposed to local rainfall (r = 0.57 @ lag = 3 months). While the downstream discharge explains a significant proportion of variability in the leading orthogonal mode of rainfall (r = 0.83), statistical relationships developed between discharge/rainfall and the distribution of aquatic primary producers confirm that both rivers (flow) and local rainfall are key optimal predictors of inundation and sites of primary producers with river flows being a better indicator of the latter. It is therefore argued that hydro-meteorological fluctuations will be key constraints on freshwater habitats and the growth of primary producers, albeit, human disturbance of flow can impact on floodplain productivity. As illustrated in this study, such constraints are reflected in the spatial patterns and changing characteristics (connectivity and spatial heterogeneity) of hot spots of primary producers and freshwater habitats (Palustrine, Lacustrine, and Riverine ecosystems). [ABSTRACT FROM AUTHOR]