Dr Chris Owers

Tropical cyclones can significantly impact mangrove forests, with some recovering rapidly, whilst others may change permanently. Inconsistent approaches to quantifying these impacts limit the capacity to identify patterns of damage and recovery across landscapes and cyclone categories. Understanding these patterns is critical as the changing frequency and intensity of cyclones and compounding effects of climate change, particularly sea-level rise, threaten mangroves and their ecosystem services. Improvements in Earth observation data, particularly satellite-based sensors and datacube environments, have enhanced capacity to classify time-series data and advanced landscape monitoring. Using the Landsat archive within Digital Earth Australia to monitor annual changes in canopy cover and extent, this study aims to quantify and classify immediate and long-term impacts of category 3-5 cyclones for mangroves in Australia. Closed canopy mangrove forests experienced the greatest immediate impact (loss of canopy cover). Most immediate impacts were minor, implying limited immediate mortality. Impacts varied spatially, reflecting proximity to exposed coastlines, cyclone track and forest structure (height, density, condition and species). Recovery was evident across all cyclones, although some areas exhibited permanent damage. Understanding the impacts and characteristics of vulnerable and resilient forests is crucial for managers tasked with protecting mangroves and their services as the climate changes.

A morphodynamic approach to coastal evolution involves recognition of internal thresholds, feedbacks and boundary conditions and should underpin coastal management. The Holocene evolution of the Bega River estuary and Tathra Beach coastal barrier was examined integrating existing sediment cores and radiocarbon dating, airborne terrestrial and marine Lidar and OSL dating. Sediment coring reveals the Bega River estuary began infilling with fluvial sand once sea levels stabilised at or near their present elevation. Radiocarbon dating suggests a prograding fluvial delta reached the coast approximately 4000¿2250 years BP. Barrier deposition commenced ~3200 years ago coinciding with the arrival of fluvial sand at the coast. Shoreline progradation of the Tathra barrier occurred at 0.15 m/year from ~3200 years to present forming a sequence of ~17 foredune ridges which were each active for an average of ~190 years. In the past ~500 years, a sand spit has restricted the entrance of the Bega River estuary to the northern end of the embayment. The infill of the Bega River estuary over the Holocene represents an internal morphodynamic threshold or tipping point, which then enabled coastal barrier deposition as fluvial sand reached the coast. The coastal system approaches another threshold as the Tathra embayment infills, and sediment may be transported northward out of the embayment. At Tathra Beach, the positive sediment budget which resulted in barrier progradation is approximately 0.55 m3/m/year. This signal is masked on the yearly to decadal scale by fluctuations in beach volume an order of magnitude greater (5¿20 m3/m/year depending on the timeframe examined). Thus longer-term datasets of beach change or reconstructions from the geological record are needed to underpin management decisions which will impact shorelines decades or centuries into the future.

National and global scale initiatives to reduce loss and promote restoration of coastal ecosystems have leveraged the capacity of mangrove and saltmarsh to contribute to climate change mitigation through carbon sequestration. The success of these programs is predicated on reliable estimates of carbon storage and how this changes over time. Efforts to describe spatial variation in below-ground carbon storage have largely focussed on surface sediments, with few studies able to characterise carbon at greater soil depths. This study demonstrates that landscape position occupied by wetland vegetation influences both carbon storage and sources, and that understanding evolutionary infill of estuaries is crucial for characterising spatial variation in carbon storage. We focussed on coastal wetlands in southeast Australia where sea level has a long history of relative stability over the past few millennia. Under these conditions, we show that carbon storage varies across three depth zones in substrate: the active root zone (associated with distribution of contemporary vegetation), inactive root zone (associated with past environmental conditions) and subtidal zone (beyond the contemporary intertidal zone). This conceptual approach relates spatial variation in carbon storage to key processes influencing carbon addition and decomposition, and can be applied elsewhere depending on the sea-level history at the specific site. We demonstrate that models that define carbon storage in the context of variation in landscape position of vegetation in the tidal frame provide improved confidence required for blue carbon assessments.

The development and refinement of methods for estimating organic carbon accumulation in biomass and soils during mangrove restoration and rehabilitation can encourage uptake of restoration projects for their ecosystem services, including those of climate change mitigation, or blue carbon. To support the development of a blue carbon method for Australia under the Emission Reduction Fund scheme we investigated; (1) whether carbon accumulation data from natural mangroves could be used to estimate carbon accumulation during restoration; (2) modeling mangrove biomass accumulation; and (3) how modeled carbon accumulation could be achieved over heterogeneous sites. First, we assessed carbon accumulation in soil and biomass pools from the global literature, finding that estimating carbon accumulation using data from natural mangroves provided similar estimates as those for restored or rehabilitated mangroves. We assessed mangrove biomass accumulation from global chronosequence studies, which we used to develop regional models for estimating biomass accumulation with restoration in Australia using values from local natural mangroves. Estimating biomass carbon accumulation using site-based means of stand biomass provided similar estimates as values estimated through use of regional means values stratified by elevation; and reduced overestimates of biomass carbon accumulation that were based on regional mean values. Modeling soil carbon accumulation over environmentally heterogeneous project sites can apply a similar approach, stratifying over variation in site elevation. Our analysis provides a strategy for modeling blue carbon pools for an Australian blue carbon method that accommodates regional differences and is based on data from natural mangroves.

Carbon mitigation services provided by coastal wetlands are not spatially homogeneous, nevertheless are commonly described on the basis of vegetation distribution within the intertidal zone. Distribution of mangrove and saltmarsh varies in response to frequency of tidal inundation, resulting in environmental gradients in edaphic factors that influence vegetation structure, and subsequently affect sedimentary carbon additions by vegetation and carbon losses by decomposition. Current sampling approaches and reporting do not adequately account for variability of carbon storage within a wetland, and assessments need to capture spatial variation associated with carbon storage to improve estimates of potential carbon mitigation services by natural ecosystems. This study quantifies the variation in near-surface carbon storage (i.e. upper 30 cm) across an intertidal gradient using a stratified sampling approach that recognises vegetation structure. Vegetation distribution and structure, as well as sedimentary controls on carbon content, explained variation in carbon storage. Saltmarsh near-surface carbon storage varied considerably between structural form. This was less evident for mangrove structural forms (i.e. tall, shrub, dwarf), which may be due to mangrove roots extending to depths beyond 30 cm. Sedimentary characteristics correlated with carbon content, demonstrating considerable influence on near-surface carbon storage within a wetland. The principal finding of this study was that variation within a wetland corresponds to the variation between sites. Stable carbon isotopes offer a means to identify previous vegetation contributions to sediment, associated with an earlier stage of wetland development, likely reflecting previous environmental conditions. A stratified sampling approach that recognises vegetation structure provides the capacity to account for variability of carbon within a wetland that is inadequately described by current sampling protocols.

This study establishes the use of the Earth Observation Data for Ecosystem Monitoring (EODESM) to generate land cover and change classifications based on the United Nations Food and Agriculture Organisation (FAO) Land Cover Classification System (LCCS) and environmental variables (EVs) available within, or accessible from, Geoscience Australia's (GA) Digital Earth Australia (DEA). Classifications representing the LCCS Level 3 taxonomy (8 categories representing semi-(natural) and/or cultivated/managed vegetation or natural or artificial bare or water bodies) were generated for two time periods and across four test sites located in the Australian states of Queensland and New South Wales. This was achieved by progressively and hierarchically combining existing time-static layers relating to (a) the extent of artificial surfaces (urban, water) and agriculture and (b) annual summaries of EVs relating to the extent of vegetation (fractional cover) and water (hydroperiod, intertidal area, mangroves) generated through DEA. More detailed classifications that integrated information on, for example, forest structure (based on vegetation cover (%) and height (m); time-static for 2009) and hydroperiod (months), were subsequently produced for each time-step. The overall accuracies of the land cover classifications were dependent upon those reported for the individual input layers, with these ranging from 80% (for cultivated, urban and artificial water) to over 95% (for hydroperiod and fractional cover). The changes identified include mangrove dieback in the southeastern Gulf of Carpentaria and reduced dam water levels and an associated expansion of vegetation in Lake Ross, Burdekin. The extent of detected changes corresponded with those observed using time-series of RapidEye data (2014 to 2016; for the Gulf of Carpentaria) and Google Earth imagery (2009¿2016 for Lake Ross). This use case demonstrates the capacity and a conceptual framework to implement EODESM within DEA and provides countries using the Open Data Cube (ODC) environment with the opportunity to routinely generate land cover maps from Landsat or Sentinel-1/2 data, at least annually, using a consistent and internationally recognised taxonomy.

Above-ground biomass represents a small yet significant contributor to carbon storage in coastal wetlands. Despite this, above-ground biomass is often poorly quantified, particularly in areas where vegetation structure is complex. Traditional methods for providing accurate estimates involve harvesting vegetation to develop mangrove allometric equations and quantify saltmarsh biomass in quadrats. However broad scale application of these methods may not capture structural variability in vegetation resulting in a loss of detail and estimates with considerable uncertainty. Terrestrial laser scanning (TLS) collects high resolution three-dimensional point clouds capable of providing detailed structural morphology of vegetation. This study demonstrates that TLS is a suitable non-destructive method for estimating biomass of structurally complex coastal wetland vegetation. We compare volumetric models, 3-D surface reconstruction and rasterised volume, and point cloud elevation histogram modelling techniques to estimate biomass. Our results show that current volumetric modelling approaches for estimating TLS-derived biomass are comparable to traditional mangrove allometrics and saltmarsh harvesting. However, volumetric modelling approaches oversimplify vegetation structure by under-utilising the large amount of structural information provided by the point cloud. The point cloud elevation histogram model presented in this study, as an alternative to volumetric modelling, utilises all of the information within the point cloud, as opposed to sub-sampling based on specific criteria. This method is simple but highly effective for both mangrove (r2 = 0.95) and saltmarsh (r2 > 0.92) vegetation. Our results provide evidence that application of TLS in coastal wetlands is an effective non-destructive method to accurately quantify biomass for structurally complex vegetation.

Carbon mitigation services provided by coastal wetlands are not spatially homogeneous. The scale of assessment at which above-ground biomass is measured will directly influence carbon storage estimates. Greater confidence in estimates is obtained with approaches that describe more variation. There is a need to improve accuracy while optimising assessment effort efficiency. Accurate quantification of carbon storage is dependent upon accurate assessment of biomass, carbon content and the extent of vegetation for which carbon storage is being assessed. This study demonstrates that vegetation structure influences above-ground biomass of mangrove and saltmarsh, resulting in considerable variability in biomass estimates and associated carbon storage of temperate coastal wetlands in southeast Australia. For mangrove, variability in above-ground biomass (Mg ha-1 ± SE) was best described by measuring height, stem diameter, crown area and vegetation density, whereby tall mangrove (3¿8 m in height; 71.50 ± 12.53 Mg ha-1) had higher biomass than both shrub (1.3¿3 m in height; 53.06 ± 6.94 Mg ha-1) and dwarf mangrove (<1.3 m in height; 10.68 ± 1.77 Mg ha-1). Saltmarsh above-ground biomass was best described by height, species and vegetation density, which demonstrated significant differences between rush saltmarsh (15.97 ± 2.35 Mg ha-1) and herbs, grasses and sedges saltmarsh (7.51 ± 0.91 Mg ha-1). The effect of this variation was compounded by carbon content (% C), which varied markedly between vegetation structural form and species (30.9¿49.8% C). Maintaining accuracy when assessing carbon storage requires mapping units that correspond to the scale of biomass assessments. Results from this study suggest that recognition of variation in biomass and carbon content of mangrove and saltmarsh vegetation structure will enhance the accuracy of estimates of carbon storage, and provide the confidence necessary for carbon storage inventories.

Coastal wetland vegetation is complex in form and function. Accurately mapping the spatial variation of vegetation complexity within these ecosystems is important for identifying areas of high conservation value that provide essential ecosystem services. In this study we delineate wetland vegetation, particularly mangrove and saltmarsh, to a vegetative morphological level that identifies spatial complexity in vegetation structure. This was achieved by integrating light detection and ranging (Lidar) and aerial imagery with an object-based approach. The results demonstrate that this is an effective methodology to identify vegetation complexity, with all study sites having greater than 90% classification accuracy. These high classification accuracies were underpinned by the use of Lidar data that provide detailed structural information about vegetation that is not captured with aerial imagery. This research highlights the importance of identifying spatial variability in vegetation structure when considering the value of coastal ecosystems and the services they provide.

Quantifying carbon storage in coastal wetland environments is important for identifying areas of high carbon sequestration value that could be targeted for conservation. This study combines remote sensing and sediment analysis to identify spatial variation in soil carbon storage for Currambene Creek, New South Wales, Australia to establish whether vegetation structure influences soil carbon storage in the upper 30 cm. Wetland vegetation was delineated to capture structural complexity within vegetation communities using Light detection and ranging (Lidar) point cloud data and aerial imagery with an object-based image analysis approach. Sediment cores were collected and analysed for soil carbon content to quantify below-ground carbon storage across the site. The total soil carbon storage in the upper 30 cm for the wetland (59.6 ha) was estimated to be 3933 ± 444 Mg C. Tall mangrove were found to have the highest total carbon storage (1420 ± 198 Mg C), however are particularly sensitive to changes in sea-level as they are positioned lowest in the intertidal frame. Conservation efforts targeted at protecting areas of high carbon sequestration, such as the tall mangrove, will lead to a greater contribution to carbon mitigation efforts.