Scale dependency of pseudo-absences selection and uncertainty in climate scenarios matter when assessing potential distribution of a rare poppy plant Meconopsis punicea Maxim. under a warming climate

Zhongrong Li,Zhongfu Zhu,Yan Wu

DOI: https://doi.org/10.1016/j.gecco.2020.e01353

IF: 4

2020-12-01

Global Ecology and Conservation

Abstract:<p>Under combined pressure from climate and land-cover change, knowledge of understanding how suitable habitat of rare species responds to changing environments both regionally and locally is in urgent need for conservation monitoring and habitat management in areas of high conservation values. This study elucidated what environmental factors determine habitat suitability of a rare 'red poppywort' <em>Meconopsis punicea</em> Maxim. (Ranunculales: Papaveraceae) on both scales and evaluated how the distribution would be affected under future climate warming. Specifically, boosted regression tree (BRT) was applied in a hierarchical way by integrating different pseudo-absence selections to identify the relative contribution of both regional climatic and local land-cover variables. An 'idealized' scenario derived from the long-term meteorological records, combined with three general circulation models under A2 and two representative concentration pathways (RCP) scenarios, was used as future climate scenarios for 2080. Here, the 'idealized' scenario was regarded as a low climate change scenario of a 2 °C increase in mean annual temperature (MAT) by 2080, compared to the RCP2.6 where MAT rise by 2.3 °C, whereas the A2 as a moderate change of a 4 °C increase and the RCP8.5 as an extreme change of a 6 °C increase. Results show that BRT models using BIOCLIM-weighted pseudo-absences provided the 'best' result for both scales according to five evaluation measures. Further, the appropriate number of pseudo-absences depended on the study extent. The habitat suitable to <em>M. punicea</em> was determined regionally by growing seasonal precipitation of 250–800 mm, an average annual temperature of −5–10 °C and an isothermal environment of 37–48%, and locally by moderate herb coverage, low density of artificial vegetation and very sparse shrubs. The distributional future predictions from an idealized scenario was more conservative than those from climate-model scenarios, but consistently exhibited a decreasing trend by 2080. This study could also inform future conservation practices and improve protection of rare plant habitat.</p>

biodiversity conservation,ecology