Do initial species communities planted affect subsequent succession? Results from a nine-year restoration experiment

Background/Question/Methods

The ecological restoration of disturbed sites may be broken down into two components: initial revegetation and subsequent succession. The Palmerton Zinc Pile Superfund Site in Palmerton, PA consists of over 2000 acres on a mountainside that experienced major soil contamination and erosion due to emissions from 82 years of zinc smelting. By the time the smelters shut down in 1980, the mountain was largely barren, but restoration efforts begun in 2003 have succeeded in revegetating over 80% of the site. These initial plantings focused on perennial C4 grasses because of their low metal uptake rates and their tolerance for shallow, metalliferous soils.

In 2012, we conducted a vegetation census, measuring percent cover of all herbaceous, shrub, and tree species in 36 subplots in each of 30 one-acre plots at the Palmerton site. These plots had been seeded in 2003 with one of three combinations of grass species. Each plot received eight perennial C4 grasses either alone, with three annual C3 grasses, or with three perennial C3 grasses. We asked whether current vegetation differs as a function of these initial plantings, in order to assess the planted species’ establishment success and potential effects on succession.

Results/Conclusions

After nine years, all perennial C4 grasses planted still occurred in virtually every plot, but no annual C3 grasses planted were detected in any plot. The perennial C3 grasses (primarily Festuca ovina or F. trachyphylla) occurred in eight of the ten plots in which they were planted, and in three plots in which they were not planted. Because these species only colonized plots within ~10 meters of plots in which they were planted, dispersal limitation is suggested.

Approximately 92 species not planted were also recorded, some of them in every plot. These most common species included the native perennial forbs Eupatorium serotinum and Ageratina altissima, the native perennial grass Agrostis perennans, and two invasive woody species, the nonnative shrub Buddleja davidii and the native tree Betula populifolia. Also of note is the annual forb Minuartia patula, which occurred in 24 of the 30 plots, is a suspected metal hyperaccumulator, and is native to the United States but found nowhere else in Pennsylvania. We also examine whether the initial species planted affected the abundance of these or other colonists, and how plant traits such as mycorrhizal association or metal accumulation strategy may relate to the successional patterns observed.