Opportunities for Integrated Ecological Analysis across Inland Australia with Standardised Data from Ausplots Rangelands

Abstract

Australian rangelands ecosystems cover 81% of the continent but are understudied and continental-scale research has been limited in part by a lack of precise data that are standardised between jurisdictions. We present a new dataset from AusPlots Rangelands that enables integrative rangelands analysis due to its geographic scope and standardised methodology. The method provides data on vegetation and soils, enabling comparison of a suite of metrics including fractional vegetation cover, basal area, and species richness, diversity, and composition. Cover estimates are robust and repeatable, allowing comparisons among environments and detection of modest change. The 442 field plots presented here span a rainfall gradient of 129–1437 mm Mean annual precipitation with varying seasonality. Vegetation measurements include vouchered vascular plant species, growth form, basal area, height, cover and substrate type from 1010 point intercepts as well as systematically recorded absences, which are useful for predictive modelling and validation of remote sensing applications. Leaf and soil samples are sampled for downstream chemical and genomic analysis. We overview the sampling of vegetation parameters and environments, applying the data to the question of how species abundance distributions (SADs) vary over climatic gradients, a key question for the influence of environmental change on ecosystem processes. We found linear relationships between SAD shape and rainfall within grassland and shrubland communities, indicating more uneven abundance in deserts and suggesting relative abundance may shift as a consequence of climate change, resulting in altered diversity and ecosystem function. The standardised data of AusPlots enables such analyses at large spatial scales, and the testing of predictions through time with longitudinal sampling. In future, the AusPlots field program will be directed towards improving coverage of space, under-represented environments, vegetation types and fauna and, increasingly, re-sampling of established plots. Providing up-to-date data access methods to enhance re-use is also a priority.

TERN, Australia’s land observatory: addressing the global challenge of forecasting ecosystem responses to climate variability and change

“…List of physical and data research infrastructure (RI) operated by TERN. changes in biodiversity are monitored continentally, and temporal changes over long timescales (5-10 years) (Tokmakoff et al 2016, Guerin et al 2017; (ii) ecosystem processes 'SuperSites' equipped with eddy covariance flux towers and from which temporal changes in ecosystem structure and function are monitored at a high level of detail in a spatially limited number of locations , Karan et al 2016, van Gorsel et al 2018; and (iii) landscapes, spatially distributed soil, environmental monitoring and remote sensing products at continental spatial scales (Grundy et al 2015, Mahoney et al 2016. TERN 2015).…”

Stocktaking the environmental coverage of a continental ecosystem observation network

“…The Ausplots field sampling protocols and datasets have been described previously (White et al 2012;Wood et al 2015;Guerin et al 2017;Sparrow et al 2019a). Here, we focus mainly on the spatial sampling of the plot network as a whole rather than the measurements taken.…”

“…Firstly, Ausplots initially focused primarily on the rangelands of the Australian inland, so it would not be surprising if wetter coastal habitats were underrepresented. Coastal areas outside the rangelands were deliberately avoided in the early sampling of Ausplots, specifically because the inland of Australia is known to be under-studied (Eyre et al 2011;Guerin et al 2017). Secondly, by maximising coverage of climatic zones with even sampling density, the stratified random strategy may have overrepresented more mesic, coastal habitats, as shown by comparison to background points (Fig.…”

“…Here, we assess the performance of a real network of monitoring sites -the Terrestrial Ecosystem Research Network's (TERN) Ausplots -in terms of representativeness by comparing its environmental coverage with that of virtual plot networks based on alternative, 'retrospective' sampling schemes. TERN Ausplots are terrestrial ecosystem monitoring sites located across Australia that have been sampled using a consistent method involving quantitative measurements of vegetation and characterisation of soils (White et al 2012;Guerin et al 2017;Cleverly et al 2019;Sparrow et al 2019a). Ausplots sites form an ecosystem surveillance monitoring network (sensu Eyre et al 2011;Sparrow et al 2019b) that now has plots at over 700 locations.…”

When macroecological transitions are a fiction of sampling: comparing herbarium records to plot‐based species inventory data

“…We compared a set of macroecological metrics calculated from the vascular plant species occurrence records of two datasets: 1) 'herbarium': herbarium records obtained via Australia's virtual herbarium that were filtered for spatial validity and certain species level identification, giving a dataset of 140 624 records of 3586 species across all habitat types (AVH 2014) and; 2) 'plots': systematically collected species inventories from field survey plots of the Biological Survey of South Australia supplemented by those from Terrestrial Ecosystem Research Network AusPlots (Dept of Environment, Water and Natural Resources 2012, Guerin et al 2017), giving a dataset of 125 085 records of 2144 species from 4740 plots, collected between 1977 and 2012. Most of the plots could be described as 'preferential' as they were neither strictly randomly located nor stratified to a sampling grid and therefore sample habitats somewhat subjectively.…”