Small-scale demographic compensation in a high-mountain endemic: the low edge stands still

Abstract

Background: Under the current warming process, with its implications of higher temperatures, less rainfall, snowfall and snow cover, low-edge populations of high-mountain plants are expected to decline. Demographic studies are useful to foresee the future dynamics of species ranges. Adaptation, phenotypic plasticity, and demographic compensation have been proposed as mechanisms to cope with climate change.

Aim: We studied the population dynamics of the high-mountain narrow endemic Armeria caespitosa at the extremes of its altitude distribution (low versus high edge), the high edge representing colonised summits that provide optimal growing conditions. The focus of the study was to establish if the species is retracting at its lower distribution range.

Methods: We used 4 years of population censuses and soil seed bank data. Population matrix models allowed us to study the stochastic population growth rates, the species long-term viability under higher frequency of extreme years, and the life cycle transitions responsible for the differences in population dynamics between elevation range edges.

Results: The low edge of the A. caespitosa elevation range appeared stable, with positive population growth rates under current environmental conditions, and a null quasi-extinction probability in the long-term under scenarios of high frequency of extremely dry years.

Conclusions: The species will likely withstand the current climate-warming scenario along its elevation range. Results supported the role of so-called demographic compensation at a small spatial scale, which we argue might be common in the Mediterranean mountains.

Keywords:

• climate change
• demographic compensation
• distribution edges
• high-mountain plants
• plant demography

Acknowledgements

We are very grateful to Dr. Luis Giménez-Benavides, and Dr. Beatriz Pías, who helped setting the sampling plots, and the staff of Parque Regional de la Cuenca Alta del Manzanares for permission to work in the protected areas. We thank Dr. Pedro Quintana-Ascencio for helping us with the stochastic growth rates handing over his R scripts. This study was funded by the Spanish Ministry of Science and Innovation's ISLAS (CGL2009‐13190-C03‐01) and LIMITES (CGL2009‐07229) research projects and by REMEDINAL2. R.G.-C. had financial support from a predoctoral fellowship from the Madrid Autonomous Government (F.P.I. Orden 3334/04 CAM) and from a postdoc contract from GLOWA Jordan River Project.