Dormancy in Caladenia: a Bayesian approach to evaluating latency

Abstract

Dormancy is common in many terrestrial orchids in southern Australia and other temperate environments. The difficulty for conservation and management when considering dormancy is ascertaining whether non-emergent plants are dormant or dead. Here we use a multi-state capture–recapture method, undertaken over several seasons, to determine the likelihood of a plant becoming dormant or dying following its annual emergent period and evaluate the frequency of the length of dormancy. We assess the transition probabilities from time series of varying lengths for the following nine terrestrial orchids in the genus Caladenia: C. amoena, C. argocalla, C. clavigera, C. elegans, C. graniticola, C. macroclavia, C. oenochila, C. rosella and C. valida from Victoria, South Australia and Western Australia. We used a Bayesian approach for estimating survivorship, dormancy and the likelihood of death from capture–recapture data. Considering all species together, the probability of surviving from one year to the next was ~86%, whereas the likelihood of observing an individual above ground in two consecutive years was ~74%. All species showed dormancy of predominantly 1 year, whereas dormancy of three or more years was extremely rare (<2%). The results have practical implications for conservation, in that (1) population sizes of Caladenia species are more easily estimated by being able to distinguish the likelihood of an unseen individual being dormant or dead, (2) population dynamics of individuals can be evaluated by using a 1–3-year dormancy period and (3) survey effort is not wasted on monitoring individuals that have not emerged for many years.