Initial effects of invasions by a forest pine (Pinus elliottii) on savanna groundcover vegetation in its introduced range and at home

Background/Question/Methods Invasive trees can cause catastrophic reductions in diversity in invaded communities. Such reductions are hypothesized to be particularly severe following introduction to a new continent. Nevertheless, invasive tree species originating from forests containing relatively shade/litter-tolerant groundcover vegetation are likely to reduce diversity of shade/litter-intolerant savanna vegetation after invasion, regardless of whether the invasion/expansion occurs in the tree’s home range or on another continent.  A relevant test of biogeographical differences in the impact of invasive forest trees therefore requires a comparison of effects of invasion on savannas both at home and in the introduced range. To our knowledge, such a comparison has not been done. We present initial results of a field study that examined the competitive effect of invasive slash pine (Pinus elliottii) from the southern US on groundcover vegetation in an invaded savanna community in Brazil (cerrado), an invaded savanna community within its native range in Mississippi (longleaf pine savanna) following fire exclusion, and its natural home community in Mississippi (slash pine maritime forest). Competitive effects were quantified by examining differences between invaded and uninvaded/restored sites and through experimental needle addition and removal in savanna and slash pine forest communities on both continents.

Results/Conclusions As a group, invaded savanna sites on both continents exhibited lower species richness and plant density than did uninvaded or restored savanna sites. Nevertheless, differences in species richness between invaded and uninvaded sites were significantly greater in Brazil than in Mississippi (94% difference in cerrado vs. 59% in longleaf pine savanna). Tree density, overhead canopy, and needle depth were significantly higher (and species richness and plant density significantly lower) at invaded sites in Brazil than at an invaded and unrestored savanna site in Mississippi or at a site containing native slash pine maritime forest. By one year after implementation of the treatments, needle addition had resulted in a highly significant reduction in both total groundcover plant density and species richness at three uninvaded cerrado sites (88% and 69%, respectively). Needle removal had weak positive effects on density and diversity at invaded sites in Brazil. Although effects of needle addition and removal remain to be examined in Mississippi (but will be presented at the meetings), initial results suggest that higher growth rates and greater needle deposition by slash pine in Brazil compared to Mississippi are in part responsible for differences in impact.