Abstract
Endemic plants in high mountains are projected to be at high risk because of climate change. Temporal demographic variation is a major factor affecting population viability because plants often occur in small, isolated populations. Because isolated populations tend to exhibit genetic differentiation, analyzing temporal demographic variation in multiple populations is required for the management of high mountain endemic species. We examined the population dynamics of an endemic plant species, Primula farinosa subsp. modesta, in four subalpine sites over six years. Stage-based transition matrices were constructed, and temporal variation in the projected population growth rate (λ) was analyzed using life table response experiments (LTREs). The variation in λ was primarily explained by the site × year interaction rather than the main effects of the site and year. The testing sites exhibited inconsistent patterns in the LTRE contributions of the vital rates to the temporal deviation of λ. However, within sites, growth or stasis had significant negative correlations with temporal λ deviation. Negative correlations among the contributions of vital rates were also detected within the two testing sites, and the removal of the correlations alleviated temporal fluctuations in λ. The response of vital rates to yearly environmental fluctuations reduced the temporal variation of λ. Such effects manifested especially at two sites where plants exhibited higher plasticity than plants at other sites. Site-specific temporal variation implies that populations of high mountain species likely exhibit asynchronous temporal changes, and multiple sites need to be evaluated for their conservation.