PARENT SESSION
Oral Session #1: Conservation Ecology: Threatened and Endangered Species. Presiding: W. Bond.
Monday, August 6, 2001. 8:00 AM to 12:15 PM. Madison Ballroom C.
Demography of the Federally listed endangered sunflower, Helianthus schweinitzii: a comparison of two models.
Workman, Michael1, Davis, Eric1,2, Estep, Bret1,3, Barden, Lawrence1, 1 2 3
ABSTRACT- Each autumn for 8 years, 1993-2000, we recorded location, height, stems, and flower heads on every individual in two populations of a Federally listed endangered perennial sunflower, Helianthus schweinitzii. We used annual transition rates from the first 5 years (4 transitions) to predict population size and 95% confidence intervals for the next 3 years, and then compared predictions with actual counts. We compared predictions of a stage-structured model (RAMAS/Stage) and an unstructured model (Dennis et al. 1991) because of the large difference in effort required to collect data appropriate to the two models (stage-structured data collection is much more labor-intensive). Both populations increased over the 4-year calibration period, 1993-1997 (lambda = 1.3 and 1.2). Both populations increased again in 1998, but both populations dropped precipitously in 1999 and 2000 (2-year declines of 44% and 62%). No model based on our calibration period could have predicted these declines. However, the unstructured model was more accurate in its predictions than the stage-structured model. Furthermore, the wider 95% confidence limits of the unstructured model were more realistic predictors of the actual decline than the confidence limits of the stage-structured model. While properly constructed stage-structured models can identify critical life stages and suggest management strategies to increase populations, in this case, an unstructured model, which required far less monitoring effort, provided more realistic predictions of trajectories of Schweinitz sunflower populations.
KEY WORDS: Helianthus schweinitzii, demography, conservation, population