Fire and population dynamics of woody plants in a neotropical savanna: matrix model projections

Ecology, June, 1999 by William A. Hoffmann

Burning increased mortality of all species, but primarily among the smallest size classes. Among adults, fire-induced mortality was much lower (Table 1). Topkill, defined as death of the aboveground stem, was much more prevalent than true mortality. Only large individuals escaped topkill, so the subshrub Periandra experienced 100% topkill, and the two shrubs experienced nearly 100% topkill. The two trees experienced considerably lower topkill among the larger size classes. Resprouting began shortly after burning.

Censuses of seed production were performed in the study plots, and in additional transects, as described by Hoffmann (1998). These transects were established in areas with other fire treatments, including biennial early dry-season burns, biennial mid dry-season burns, biennial late dry-season burns, and quadrennial mid dry-season burns. Seedling establishment data were obtained from a series of experimental plots in which seeds were placed, as is presented by Hoffmann (1996a). Additional establishment data collected in an identical manner were since obtained for Periandra and Myrsine (Hoffmann, unpublished data).

These studies revealed that sexual reproduction is unsuccessful in years that burning occurs, because fire destroys seeds, flower buds, or developing fruit. In subsequent years, size-specific seed production of Periandra is the equal to the preburn level. For Miconia, few fruit are produced in the first year after burning, but, in the second and third years after burning, size-specific seed production is greater than the preburn level. For Rourea, size-specific seed production is greatest in the first year after burning, and decreases in subsequent years. For both Myrsine and Roupala, seed production is low in the first year after burning, and gradually increases to preburn levels in subsequent years.

Rourea, Myrsine, and Roupala reproduce vegetatively by producing root suckers (Hoffmann 1998). These root suckers are defined as new individuals originating from root buds at some distance from the parent stem. Hoffmann (1998) describes estimation of sucker production. Following the method of Ribbens et al. (1994), I used a maximum-likelihood estimator to fit the allometric relationship between stem diameter and sucker production, within the study plots. Fire significantly stimulated the production of root suckers by all three species. Suckers have lower rates of fire-induced mortality than do seedlings of similar age (Table 1; Hoffmann 1998).

For estimating demographic parameters, an individual was defined as a stem, or a group of stems, connected at the root crown. Stems not connected at the root crown were considered separate individuals. According to this definition, root suckers were not considered part of the parent individual, so there may have been physiological connections between stems considered to be separate individuals. However, for Myrsine and Roupala, the connection between parent and offspring typically disintegrated within one year. For individuals with more than one stem connected at the root crown, individual size was taken to be the diameter of the largest stem.