Data Paper. Data Paper
Florida scrub supports many endemic species in a pyrogenic shrubland on xeric yellow sands. One genus with multiple endemics is the mint Dicerandra, with its five narrowly endemic species most prevalent in edges and recently burned areas. Study of the demography of Dicerandra frutescens, an endemic and endangered mint restricted to Florida scrub, began in September 1988. Eleven populations have been studied at various locations across Archbold Biological Station and at one privately held site. Some populations, particularly along fire lanes, have been added during the study. Survival and recruitment data were initially collected monthly, but were later collected quarterly in September, December, March, and June. Each annual census in September also recorded the life history stage for that "year" (the year including any censuses after the prior September census) and measures of size (basal diameter, number of branch tips, number of reproductive branch tips). Survival variables representing monthly, quarterly, and annual survival were also assembled. These survival variables contain information on the type of plant recruiting (seedling or new adult) or, for plants already in the data set, whether the plant survived the interval, died during the interval, was already (previously) dead, or whether survival could not be ascertained (no tag, no plant).
Several populations were subjected to prescribed burns during the study. For years with burns, variables summarize whether plants (or quadrats) were burned or not. This data set consists of 5530 records, each an individual plant, from 1988 through 2004. Published analyses based on data from 1990–2000 found that finite rates of increase were greater than 1 for populations less than six years post-fire, but thereafter populations were predicted to decline. Similarly, populations in fire lanes were most successful when the time-since-disking was short. A yard-edge population showed variable demography. Stochastic simulations in Florida scrub sites suggested an optimal fire return interval of 6–12 years (regular fires) or 6–21 years (stochastic fires).
Data collection on this project is continuing, and we anticipate updating data periodically.
Key words: fire effects on demography; fire frequency; fire lanes; fire management; long-term data set; population viability analysis; seedling cohorts; stochastic population modeling, survival; transition matrix.