Does removal of Gypsophila paniculata (baby’s breath) from Lake Michigan sand dunes restore native community structure and ecosystem function?

Background/Question/Methods

Gypsophila paniculata (baby’s breath) is a serious plant invader of Lake Michigan sand dune systems.  On some dunes, G. paniculata comprises 80% of all vegetation.  It is widely believed that G. paniculata creates problems for the dunes by over-stabilizing habitat, making it unsuitable for native plant species such as the federally-threatened Cirsium pitcheri (Pitcher’s thistle), and may be capable of out-competing other native species due to its deep root system.  Due to growing concern about the impacts of G. paniculata, The Nature Conservancy in Michigan and the National Park Service are actively managing dune systems with the goal of G. paniculata eradication.  Since 2008, field crews have removed G. paniculata manually from over 450 acres of dune habitat.  We evaluated the effectiveness of these restoration efforts by monitoring a variety of community structure and ecosystem function measures in 15 sites in northwest lower Michigan in the three years following treatment.

Results/Conclusions

Management efforts to date have been very successful in reducing G. paniculata populations to less than 10% cover.  G. paniculata removal has not altered plant species richness or success of other non-native species.  Native plant cover is increasing, with managed sites now more closely resembling non-invaded sites, though C. pitcheri abundance has not changed. Unmanaged invaded areas have more total insect numbers, more pollinators, and more herbivores than areas where G. paniculata has been removed.  Managed areas have increased exposed ground, though it is not clear whether this translates to increased sand movement, an important component of native dune habitat which can facilitate colonization of many native species. Soil mycorrhizal communities are not affected by G. paniculata invasion or management efforts, though some soil nutrients (P, Ca) seem linked to G. paniculata invasion and are not responding to management efforts.  Overall, our results suggest that removal of G. paniculata has had only minimal impacts on plant community structure and ecosystem function three years after initial management, but increases in native plant cover and lack of other non-natives taking advantage of the open niche left by G. paniculata offer hope for longer term restoration goals.