Estimation of population size and probabilities of survival and detection in a population of Mead’s milkweed

This is the publisher's version, also available electronically from http://www.esajournals.org/doi/abs/10.1890/0012-9658%282003%29084%5B0791%3AEOPSAP%5D2.0.CO%3B2.Statistical similarities exist between estimating numbers of cryptic animals and of inconspicuous plants. Without flowers, Mead's milkweed (Asclepias meadii) is an unobtrusive prairie plant, and most plants flower irregularly. We used maximum likelihood to estimate probabilities of survival and detection for milkweeds in annual censuses during 1992–1999. Initially, we assumed no recruitment and fit models to all data and to flowering plants only. Because plants were marked when located, probabilities of resighting exceeded those of initial discovery. Plants were most likely to flower and be detected in years when the prairie was burned. We marked 177 plants in eight years but estimated the 1992 population to be 337 or 191 plants with 166 or 121 surviving to 1999, depending on the data set. Thus, estimated population size exceeded number of plants seen. Estimated annual survival probability was generally ≥0.95, but aggregate survival over eight years predicted rapid extinction without recruitment. When we included recruitment, estimates of survival changed little, and estimated population size varied between 118 and 147 individuals. Discovery of new plants in two additional years (2000 and 2001) appeared to be consistent with required recruitment, but simple counts of these plants did not track population trends