Effects of habitat deterioration on population dynamics and extinction risk of an endangered, long-lived perennial herb (Scorzonera humilis)

1 Habitat deterioration has resulted in the extinction of many local populations of plants. However, in long-lived plants there may be a time lag between habitat deterioration and a decline in population size, although some vital rates may be immediately affected. 2 We studied the dynamics and viability of populations of the endangered, long-lived perennial Scorzonera humilis to different levels of agricultural intensification in five populations over a period of 4 years at the genet level and for 3 years at the rosette (ramet) level. 3 Asymptotic growth rates varied little between years and the observed stage structure was very similar to the stable stage structure. Mortality of adult genets of S. humilis was low and their life expectancy was several decades. By contrast, turnover of ramets was high and their life expectancy was less than 4 years. 4 Calculations based on mean transition matrices indicated that populations at nutrient-poor sites would grow (lambda > 1), whereas those at nutrient-rich sites would continuously decline (lambda < 1) if conditions stayed the same. Life-table response experiment analysis revealed that the lower growth rate of nutrient-rich populations was primarily due to a lower survival of large plants, but also due to lower establishment and growth of new plants. 5 Stochastic simulations indicated that the extinction risk for medium-sized populations at nutrient-poor sites is very low, whereas the long-term survival even of large extant populations at nutrient-rich sites is unlikely unless favourable conditions for the establishment of young plants are restored. 6 Fertilization did not influence survival, but an additional cutting early in the season significantly reduced survival of S. humilis. Stochastic simulations indicated that a combination of fertilizing and early cutting drives populations quickly to extinction and that the long-term survival of late-cut, NPK-fertilized populations is also unlikely. 7 The results suggest that in long-lived plants, population size may not be a suitable predictor of long-term viability of populations because of time delays in their response to habitat deterioration. A detailed demographic analysis may therefore be necessary to assess realistically the conservation status of long-lived perennials.