POPULATION-GROWTH AND VIABILITY ANALYSES OF THE CLONAL WOODLAND HERB,ASARUM CANADENSE

1 Long-term demographic field studies combined with simulation models that incorporated two types of stochasticity were used to investigate the relative impact of reproduction and survival on the persistence of populations of the North American woodland herb, Asarum canadense. 2 Field data were collected over 7 years in replicate plots located in e arly and late successional forest habitat. By following marked ramets from year to year it proved possible to investigate the demography bot h of ramets and of groups of genetically identical ramets (clones). 3 A. canadense exhibited considerable temporal and spatial variation in reproductive success, survivorship, and population growth rates. Popul ations of ramets and clones in late successional forest habitats held their own, while those in early successional forest habitats declined. 4 Sexual and clonal reproduction had less impact on population growth than did survival. Nonetheless, seedling recruitment was important, a s indicated by the relatively high rates of genet turnover in A. canad ense populations. 5 Results from simulations indicated that the vast m ajority of genets in both habitat types failed to reproduce. Over thei r lifetime, simulated genets produced a maximum of 44 and 77 seedlings in early and late successional forest habitat, respectively. 6 Althou gh A. canadense genets are potentially immortal, the maximum life span for 2000 simulated genets was 104 years. Simulated genets lived a mea n of 3.0 and 4.8 years in early and late successional forest habitat, respectively. 7 Simulated A. canadense populations only persisted in l ate successional forest habitat. Within simulated late successional fo rest populations, the number of ramets tended to decrease over time, w hile the number of genets tended to increase over time. 8 The minimum viable population size for A. canadense ramets in late successional fo rest was 25 individuals with environmental stochasticity only, and 100 0 individuals with both environmental and demographic stochasticity; f or genets in late successional forest, the corresponding numbers were 20 and 25, respectively. 9 Ramet and genet dynamics differed greatly i n A. canadense, highlighting the importance of long-term demographic s tudies at each of these levels.