Elsevier

Ecological Indicators

Volume 117, October 2020, 106616
Ecological Indicators

Predicting hot spots of aquatic plant biomass in a large floodplain river catchment in the Australian wet-dry tropics

https://doi.org/10.1016/j.ecolind.2020.106616Get rights and content

Abstract

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.

Introduction

Numerous ecological and socio-economic functions supported by wetlands and freshwater habitats such as regulating nutrient cycle, maintaining fishery production and the role of aquatic plants in enhancing water clarity, among other functions have been widely reported (e.g., Chen et al., 2014, Zhao et al., 2012, Tockner et al., 2010, Midwood and Chow-Fraser, 2010, Keddy et al., 2009, Gidley, 2009, Cazzanelli et al., 2008, Ozesmi and Bauer, 2002). The large-scale assessment of wetlands and floodplains is therefore required to improve our contemporary understanding of wetland functions and their response to the degrading impacts of anthropogenic actions and strong fluctuations in climate. This is particularly the case for floodplains that are currently unaltered but under pressure from development that would impact their ecological function (Kingsford, 2000). However, physical access to several wetlands and freshwater habitats can be restricted due to difficult topography. Inaccessibility during times of inundation and extreme floods and technological limitations associated with field surveys are other key factors limiting the large-scale assessments of floodplain wetlands. While inaccessibility makes in-situ sampling challenging, aquatic primary producers such as macrophytes often thrive in freshwater habitats and floodplains that are extremely difficult to reach and perhaps unsafe (e.g., Zhao et al., 2012, Davranche et al., 2010), thus making traditional field surveys not only expensive but nearly impossible and risky. The ongoing development of satellite technologies are helping to overcome these challenges.

Several improvements in satellite remote sensing optical systems that have occurred in recent years have resulted in increased use of these technologies for the assessment of floodplain inundation, inventorying of wetlands, mapping the distribution of floods and aquatic plants, assessing planform changes, monitoring and extraction of surface water in extensive floodplains are growing (see, e.g., Tulbure and Broich, 2019, Normandin et al., 2018, Khandelwal et al., 2017, Dewan et al., 2017, Ward et al., 2016, Feyisa et al., 2014, Syvitski et al., 2012, Tockner et al., 2010, Sakamoto et al., 2007, Midwood and Chow-Fraser, 2010, Ozesmi and Bauer, 2002). The general isolation of the Australian wet-dry tropics, coupled with increasing pressures for water resource development on floodplain ecosystems and subsequent threats to ecological assets (e.g., Ward et al., 2014) has necessitated the use of remote sensing methods to assess floodplains on a large-scale. Such assessment is vital, not only to improve contemporary knowledge on drivers of inundation and aquatic primary production but to highlight the critical water needs for ecosystem health that will foster relevant policy and aid sustainability.

Tropical floodplain ecosystems are typically characterised by strong seasonality in rainfall, which in most cases results in considerable river flows and floodplain inundation during the wet seasons (e.g., Ward et al., 2014, Ward et al., 2013). Consistent with this pattern, the Flinders River catchment in the wet-dry tropics of northern Australia, experiences large fluctuations in rainfall resulting in long dry spells and seasonal flooding. Characterizing their seasonal spatial patterns and variability is critical to assess the water requirements for primary production and aquatic food webs. Whereas floodplain inundation mapping has been undertaken for some river basins in Australia, though at much smaller scales (e.g., Tulbure et al., 2016, Fisher et al., 2016, Ward et al., 2013, Ward et al., 2014), the characteristics (e.g., duration and extent) of large-scale seasonal inundation for the Flinders catchment, especially during major summer floods are largely undocumented.

These characteristics underpin knowledge on the impacts of climate change and/or anthropogenic influence on the distribution of primary production in aquatic ecosystems within wet-dry tropical systems. Isolated and discreet water holes within these systems have been identified as highly productive centers in terms of phytoplankton biomass production, macrophyte, periphyton and other primary producers that generate biomass for secondary aquatic producers (e.g., Burford et al., 2016, Ward et al., 2016, Waltham et al., 2013, Faggotter et al., 2013). While inundation extent in saline supratidal mudflats in wet-dry tropical regions of Australia was found to be an important indicator for the rates of primary productivity (see, Burford et al., 2016), the vulnerability of freshwater systems and aquatic habitats to climate influence and changes in human-water policies and management are strong motivations warranting large-scale assessments. In addition to this, the variability in reflectance values of aquatic vegetation caused by poor discrimination of aquatic plants from water signals in Lakes, especially during flood periods, and the limitations of the near-infrared wavelength in aquatic plant biomass prediction (e.g., Chen et al., 2018, Davranche et al., 2010, Cho et al., 2008, Silva et al., 2008, Malthus and George, 1997) have been reported. While these are pertinent issues, necessitating further research, the aforementioned reports were site-specific studies conducted within small Lakes, reservoirs, and wetlands/catchments.

On a global scale, the relationship between biophysical catchment properties (e.g., rainfall seasonality, dryness index, and topographic slope) and streamflow characteristics have been identified (Beck et al., 2013). In a study that investigated how drivers of streamflow characteristics vary at water management and continental scales in eastern Australia, dryness index, the fraction of photosynthetically active radiation from vegetation and soil properties accounted for observed variability in streamflow characteristics (Trancoso et al., 2017). Given the importance of floodplain rivers in providing essential services for society and ecosystems such as water supply, production of hydropower, flood control among others (e.g., Kennard et al., 2010, Keddy et al., 2009), process-based knowledge of the cascading impacts of hydro-meteorological fluctuations on ecological processes is crucial. This is because, among other reasons, they can directly feed into management and policy frameworks for biodiversity conservation and ecological assessments (Klein et al., 2009). Even though some tropical wetlands and catchments along the Australian Gulf rivers have being studied in recent times (e.g., Ndehedehe et al., 2020, Ward et al., 2016, Ward et al., 2014), the ability to answer key hydrological and ecological questions is still lacking and could be limited due to dissimilarity in biophysical catchment properties (e.g., soil properties, rainfall variability, evaporation rates, and predominant land cover states) on different spatial scales. Some of these questions include, what is the link between total downstream floodplain inundation patterns and flow regimes? What are the key drivers of hot spots of floodplain inundation and distribution of aquatic plants biomass along the Australian wet-dry tropics?

It is true that limited water availability resulting from large hydro-climatic fluctuations and decadal-scale droughts alters the biophysical properties of catchments in largely water deficit regions. This leads to non-stationary behaviour in the response of hydrological stores to rainfall as reported in some recent studies (e.g., Deb et al., 2019, Ndehedehe et al., 2019). In typical arid areas characterised by complex water use mechanisms during the dry season and where surface vegetation are morphologically and physiologically adapted to such environments (Guan et al., 2015, Seghieri et al., 2012, Huber et al., 2011), novel hydrological indicators to support the assessment of climatic constraints on agro-ecosystems have been explored (Ndehedehe et al., 2018). Therefore, the aforementioned questions are important and required to establish potential ecological problems that may result from considerable hydro-meteorological fluctuations (e.g., flood and drought events) and or change in government water policies or freshwater harvesting. With suitable analytical methodologies and empirical frameworks, these questions can be articulated and used as important tools in communicating environmental ecosystem health targeted at optimizing management initiatives, goals, and perspectives (Logan et al., 2019). To improve understanding on the connection between ecology and wetland hydrology of the Australian wet-dry tropics, an empirical framework based on quantitative assessment of relevant indicators is therefore required. While providing a basis for the use of remote sensing technique for large-scale assessments of remote floodplain systems, such a framework will also provide decision makers and other stakeholders with a quantitative basis for predicting the impacts of water resource development on floodplain wetlands.

The main aim of this study is to assess changes in large-scale seasonal total inundation and predict the spatial distribution of hot spots of primary producers (aquatic plant biomass) in a large floodplain wetland in the wet-dry tropics of Australia. The specific objectives are to (i) assess large-scale spatio-temporal changes in total floodplain inundation and (ii) predict the spatial distribution of hot spots of primary producers by integrating remote sensing indicators with digital elevation model in a classification tree model. The relationships of the predicted hot spots in (ii) above with total downstream floodplain inundation, rainfall and discharge are explored to understand the connection between ecology and wetland hydrology. Notably, remote sensing based inundation metrics have been used in quantifying floods and mapping floodplain inundation. But the need to assess the stability of these indicators in quantifying open water features and total floodplain inundation on complex landscapes such as the Flinders catchment is essential to provide more knowledge on the role of climate variability on changes in wetland hydrology. While this has also been undertaken in this study, an empirical framework that integrates a multivariate analysis of hydrological indicators (e.g., river flow) is developed for the first time and used for the quantitative assessment of local wetland hydrology. Another key application of this framework in this study is to assess the influence of inter-annual variability of rainfall on total floodplain inundation and spatial distribution of aquatic plant biomass accumulation. Such assessment is crucial and provides insight that underpins the prioritization of future surface water developments that are currently under consideration,1 conservation planning, and wetland management in this region.

Section snippets

Study area

The study region is one of the southern gulf catchments in semi-arid tropical climate of northern Australia, the Flinders catchment (Fig. 1), and has an approximate area of 109,000 km2 (e.g., Waltham et al., 2013). The unique geological features, topography, landscape characteristics and hydrology of the funnel-shaped Flinders catchment (Fig. 1a) are interesting and, includes valuable resources, e.g., minerals, coal, groundwater aquifers and soil, among others. Apart from its rich geological

Delineation of inundated areas and floodplain inundation mapping

This section is designed to deliver spatial information on floodplain inundation dynamics through a comparative analyses of surface water extraction metrics using 2009 summer flood as a case study. To this end, the major floods in the wet season of 2009 was assessed using three different but prominent water extraction metrics (MNDWI, AWEInsh and AWEInsh) derived from multi-band Landsat imagery. Moreover, in addition to assessing the hydrological indicators of floodplain productivity, the recent

Discussion

Given the role of climate and human activities as key indicators of processes that may affect water availability and the productivity of floodplain ecosystems, understanding such implications are important for environmental water allocation and the ecological communities supported by the wetlands and freshwater systems that generate multiple strings of economic values and services. Apart from the explicit representation of seasonal floodplain inundation dynamics, an ideal satellite-derived

Conclusion

Given the role of climate and human activities as key indicators of processes that may affect water availability and the productivity of floodplains ecosystems, understanding such implications is important for environmental water resource allocation and the ecological communities supported by the wetlands and freshwater systems that generate multiple strings of economic values and services. These values have generated increasing interest in water resource development among all levels of

CRediT authorship contribution statement

Christopher E. Ndehedehe: Writing - original draft, Writing - review & editing, Methodology, Formal analysis, Conceptualization, Investigation, Software, Data curation. Ben Stewart-Koster: Conceptualization, Investigation, Writing - review & editing. Michele. A. Burford: Conceptualization, Funding acquisition, Project administration. Stuart E. Bunn: Conceptualization, Project administration, Funding acquisition.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This work is supported with funding from the Australian Government’s National Environmental Science Program. The Landsat data used in this study were retrieved from USGS data portal. The authors are grateful to the Queensland Government for the river discharge observations and precipitation data accessed from the SILO climate database. The invaluable comments of two anonymous reviewers, which helped improved the quality of this manuscript are greatly appreciated.

References (92)

  • A. Fisher et al.

    Comparing landsat water index methods for automated water classification in eastern Australia

    Remote Sens. Environ.

    (2016)
  • B.-C. Gao

    NDWI’a normalized difference water index for remote sensing of vegetation liquid water from space

    Remote Sens. Environ.

    (1996)
  • A. Getirana et al.

    Hydropower dam operation strongly controls Lake Victoria’s freshwater storage variability

    Sci. Total Environ.

    (2020)
  • S. Huber et al.

    Water availability as the driver of vegetation dynamics in theAfrican Sahelfrom 1982 to 2007

    Global Planet. Change

    (2011)
  • A. Khandelwal et al.

    An approach for global monitoring of surface water extent variations in reservoirs using modis data

    Remote Sens. Environ.

    (2017)
  • T. Malthus et al.

    Airborne remote sensing of macrophytes in Cefni Reservoir, Anglesey, UK

    Aquatic Botany

    (1997)
  • M. Masocha et al.

    Surface water bodies mapping in Zimbabwe using landsat 8 OLI multispectralimagery: A comparison of multiple water indices

    Phys. Chem. Earth, Parts A/B/C

    (2018)
  • M. Montazerolghaem et al.

    Long-term variability of the leading seasonal modes of rainfall in south-eastern Australia

    Weather Climate Extremes

    (2016)
  • C. Ndehedehe et al.

    Understanding changes in terrestrial water storage over West Africa between 2002 and 2014

    Adv. Water Resour.

    (2016)
  • C.E. Ndehedehe et al.

    Is terrestrial water storage a useful indicator in assessing the impacts of climate variability on crop yield in semi-arid ecosystems?

    Ecol. Ind.

    (2018)
  • C.E. Ndehedehe et al.

    Spatio-temporal variability of droughts and terrestrial water storage over Lake Chad Basin using independent component analysis

    J. Hydrol.

    (2016)
  • C.E. Ndehedehe et al.

    Modelling the impacts of global multi-scale climatic drivers on hydro-climatic extremes (1901–2014) over the Congo basin

    Sci. Total Environ.

    (2019)
  • C.E. Ndehedehe et al.

    Changes in hydro-meteorological conditions over tropical West Africa (1980-2015) and links to global climate

    Global Planet. Change

    (2018)
  • C.E. Ndehedehe et al.

    Satellite-derived changes in floodplain productivity and freshwater habitats in northern Australia (1991–2019)

    Ecol. Ind.

    (2020)
  • C.E. Ndehedehe et al.

    Assessing land water storage dynamics over Southern America

    J. Hydrol.

    (2020)
  • C.E. Ndehedehe et al.

    Hydrological controls on surface vegetation dynamics over West and Central Africa

    Ecol. Ind.

    (2019)
  • G. Rondeaux et al.

    Optimization of soil-adjusted vegetation indices

    Remote Sens. Environ.

    (1996)
  • T. Sakamoto et al.

    Detecting temporal changes in the extent of annual flooding within theCambodia and the Vietnamese Mekong Delta from MODIStime-series imagery

    Remote Sens. Environ.

    (2007)
  • J.P. Syvitski et al.

    Floods, floodplains, delta plains - a satellite imaging approach

    Sed. Geol.

    (2012)
  • M.G. Tulbure et al.

    Spatiotemporal patterns and effects of climate and land use on surface water extent dynamics in a dryland region with three decades of Landsat satellite data

    Sci. Total Environ.

    (2019)
  • M.G. Tulbure et al.

    Surface water extent dynamics from three decades of seasonally continuous landsat time series at subcontinental scale in a semi-arid region

    Remote Sens. Environ.

    (2016)
  • D. Ward et al.

    Floodplain inundation and vegetation dynamics in the alligator rivers region (kakadu) of northern australia assessed using optical and radar remote sensing

    Remote Sens. Environ.

    (2014)
  • P. Wolski et al.

    Keeping it simple: Monitoring flood extent in large data-poor wetlands using MODIS SWIR data

    Int. J. Appl. Earth Obs. Geoinf.

    (2017)
  • J. Worden et al.

    Surface water detection in the Caucasus

    Int. J. Appl. Earth Obs. Geoinf.

    (2020)
  • D. Zhao et al.

    Remote sensing of aquatic vegetation distribution in Taihu Lake using an improved classification tree with modified thresholds

    J. Environ. Manage.

    (2012)
  • Z. Zhu et al.

    Improvement and expansion of the fmask algorithm: cloud, cloud shadow, and snow detection for landsats 4–7, 8, and sentinel 2 images

    Remote Sens. Environ.

    (2015)
  • S.R. Balcombe et al.

    Temporal changes in fish abundance in response to hydrological variability in a dryland floodplain river

    Marine Freshwater Res.

    (2009)
  • F.A. Bazzaz

    The response of natural ecosystems to the rising global CO2 levels

    Annu. Rev. Ecol. Syst.

    (1990)
  • H.E. Beck et al.

    Global patterns in base flow index and recession based on streamflow observations from 3394 catchments

    Water Resour. Res.

    (2013)
  • Bernstein, L.S., Adler-Golden, S.M., Jin, X., Gregor, B., Sundberg, R.L., 2012. Quick atmospheric correction (quac)...
  • Bernstein, L.S., Adler-Golden, S.M., Sundberg, R.L., Levine, R.Y., Perkins, T.C., Berk, A., Ratkowski, A.J., Felde, G.,...
  • S. Bunn et al.

    Tropical Floodplain Food Webs – Connectivity and Hotspots

    (2015)
  • S.E. Bunn et al.

    Sources of organic carbon supporting the food web of an arid zone floodplain river

    Freshw. Biol.

    (2003)
  • S.E. Bunn et al.

    Flow variability in dryland rivers: boom, bust and the bits in between

    River Res. Appl.

    (2006)
  • M.A. Burford et al.

    Inundation of saline supratidal mudflats provides an important source of carbon and nutrients in an aquatic system

    Mar. Ecol. Prog. Ser.

    (2016)
  • M. Cazzanelli et al.

    Emergent and floating-leaved macrophytes as refuge for zooplankton in a eutrophic temperate lake without submerged vegetation

    Hydrobiologia

    (2008)
  • Cited by (0)

    View full text