Spatial variation in the ongoing and widespread decline of a keystone plant species

Abstract

Extensive dieback in dominant plant species in response to climate change is increasingly common. Climatic conditions and related variables, such as evapotranspiration, vary in response to topographical complexity. This complexity plays an important role in the provision of climate refugia. In 2008/2009, an island-wide dieback event of the keystone cushion plant Azorella macquariensis Orchard (Apiaceae) occurred on sub-Antarctic Macquarie Island. This signalled the start of a potential regime shift, suggested to be driven by increasing vapour pressure deficit. Eight years later, we quantified cover and dieback across the range of putative microclimates to which the species is exposed, with the aim of explaining dieback patterns. We test for the influence of evapotranspiration using a suite of topographic proxies and other variables as proposed drivers of change. We found higher cover and lower dieback towards the south of the island. The high spatial variation in A. macquariensis populations was best explained by latitude, likely a proxy for macroscale climate gradients and geology. Dieback was best explained by A. macquariensis cover and latitude, increasing with cover and towards the north of the island. The effect sizes of terrain variables that influence evapotranspiration rates were small. Island-wide dieback remains conspicuous. Comparison between a subset of sites and historical data revealed a reduction of cover in the north and central regions of the island, and a shift south in the most active areas of dieback. Dieback remained comparatively low in the south. The presence of seedlings was independent of dieback. This study provides an empirical baseline for spatial variation in the cover and condition of A. macquariensis, both key variables for monitoring condition and 'cover-debt' in this critically endangered endemic plant species. These findings have broader implications for understanding the responses of fellfield ecosystems and other Azorella species across the sub-Antarctic under future climates.

Widespread dieback in a foundation species on a sub‐Antarctic World Heritage Island: Fine‐scale patterns and likely drivers

“…The survey coincided with the active summer growing season of A. macquariensis. The site locations were determined using a random stratified survey design using four terrain classes and eight geographic units to ensure that the diversity of microclimates were surveyed across the island (method described in detail in Dickson et al 2019). All sites contained A. macquariensis.…”

“…Latitude was hypothesised to be a proxy for macro-climate by Dickson et al (2019), although latitude is also correlated with elevation (r = 0.31). Elevation has a lapse rate of À0.71°C 100 m -1 on the island, which increases non-linearly (Tweedie & Bergstrom 2000).…”

“…Azorella macquariensis takes up most water through fine adventitious roots within the canopy (Bergstrom et al 2015), sourcing water from rainfall and low cloud that condenses on the leaves. Therefore, other potential proxies for atmospheric water, such as cloud cover, cloud height and wind direction (moisture), were considered; however, they would necessarily be topographically derived and subject to the limitations identified in Dickson et al (2019). Soil moisture is an important determinant of plant water stress, which we partially account for by incorporating gravel content in the models as a proxy for water holding capacity.…”

“…using topography as a proxy for local climate conditions) of A. macquariensis dieback showed latitude (N-S location on island) to be a significant predictor of plant condition. However, it shed little additional light on the possible microscale conditions associated with spatially highly variable patterns of plant decline (Dickson et al 2019).…”

“…Alternatively, if dieback severity is predicted by pathogen activity then dieback is expected to be high at sites with fewer, less extreme cold days (Burgess et al 2017) and high humidity (Huber & Gillespie 1992). Building on the coarse-scale climate and cushion condition approach taken by Dickson et al (2019), and here, we use microclimate and fine-scale cushion condition classes to further tease apart these putative drivers of dieback. The implications of the results are discussed in the context of the future persistence and management of this critically endangered foundation species under projected climate change.…”