Linking vegetative dormancy to fitness in two long‐lived herbaceous perennials

Vegetative dormancy occurs in many plant families, but its evolutionary context remains a mystery. We asked whether vegetative dormancy is an adaptive response to environmental stress and environmental stochasticity in certain long‐lived plant species. We conducted an in situ experimental study in two and three populations of Cypripedium calceolus and Cephalanthera longifolia , respectively, in Estonia. Plants were defoliated, shaded, or simply observed at the beginning of the growing season in 2002 and 2003, and monitored demographically through 2008. We assessed links between fitness and vegetative dormancy using stochastic life table response experiments (SLTREs), in which the impact of treatment on the log stochastic population growth rate ( a = log λ S ) via shifts in projection matrix transitions in treated plants relative to controls was used to assess the fitness impacts of treatment‐induced life history responses. In Cypripedium , the observed lifespans of individuals that became vegetatively dormant in 2003/04 was significantly higher than plants that had not done so ( P = 0.050). Defoliation and shading resulted in lower levels of flowering in both species. Both defoliation and shading decreased a relative to controls in Cypripedium and Cephalanthera . Defoliation‐ and shading‐induced shifts in transitions involving vegetative dormancy were generally associated with significantly positive SLTRE contributions to Δ a , and shifts in the standard deviations of demographic rates generally contributed little to Δ a . Thus, vegetative dormancy is likely to be an adaptive response to environmental stress and stochasticity. Further work on the genetic basis to vegetative dormancy will clarify whether enough heritability may have existed in the past, or exists now, to support vegetative dormancy as an adaptation.