Rhizome removal and clipping of invasive hay-scented fern Dennstaedtia punctilobula to encourage understorey regeneration in a pine plantation at Quabbin Reservation, Massachusetts, USA
Published source details
de la Cretaz A.L. & Kelty M.J. (1999) Establishment and control of hay-scented fern: a native invasive species. Biological Invasions, 1, 223-236
Published source details de la Cretaz A.L. & Kelty M.J. (1999) Establishment and control of hay-scented fern: a native invasive species. Biological Invasions, 1, 223-236
Hay-scented fern Dennstaedtia punctilobula is a native woodland understory species in the USA. It generally occurs at low density and does not noticeably interfere with the growth of other understory plants. However, it can be invasive under certain conditions e.g. increased light availability following tree thinning, and grazing and browsing of competing plants by herbivores, e.g. white-tailed deer Odocoileus virginianus deer, may result in development of a dense, nearly monospecific fern understory that inhibits tree seedlings and growth of other plants.
Study site: The study was conducted in a thinned red pine Pinus resinosa -white pine P.strobus plantation with nearly 100% fern ground cover in the Quabbin Reservation, Massachusetts.
Experimental design: Four blocks containing 12 (2 x 2 m) plots each were established. Each plot was assigned one of four fern control treatments and one of two birch seed treatments (sown; unsown). Each fern treatment was replicated three times in each block; two unsown and one unsown. The fern treatments were:
i) Removal - rhizome mat and associated litter removed in autumn 1995 (after frond die-back); exposed mineral soil broken up with a garden fork to 2-5cm depth, mimicking scalping or root raking management;
ii) Mixing - in autumn 1995, rhizome mat and litter removed and placed on a tarpaulin; soil broken up to 2-5 cm depth with a garden fork; rhizome material broken up and torn into smaller parts, and these and removed litter replaced on the plot and mixed with the soil. The treatment mimicked the management technique of scarification;
iii) Clipping - fern fronds clipped weekly in 1996 and 1997 from spring until the end of the growing season. This removed the fern canopy and mimicked mowing;
iv) Control - no treatment.
The birch seed treatments were:
i) Black birch Betula lenta seed sown at a rate of 1 g/mÂ² (approx. 1,700 seeds/mÂ²) on the soil surface in autumn 1995;
ii) Control - unsown.
Within quadrats in each plot at the end of the growing seasons of 1996 and 1997, woody seedlings were identified and counted; height of the tallest seedling of each species within each grid square was measured (to calculate average tallest height/plot). Herbaceous species that germinated were identified and presence in each grid square was recorded. Percent cover was estimated for grasses and sedges. Ferns were identified, counted, and lengths of a sample of fronds measured.
Rate of fern regrowth after a year since clipping ceased was measured in 12 clipping treatment plots.
Removal: After the two growing seasons, rhizome and litter removal many woody and herbaceous plants germinated and grew.
Mixing: There was an initial germination response but poor seedling survival as the fern canopy regrew to near pre-treatment density;
Clipping: Repeateded clipping over two growing seasons resulted in a lower germination than the removal treatment, but rapid growth of woody seedlings.
Conclusions: The hay-scented fern canopy and the rhizome/root mat inhibited tree seedling establishment. Black birch, whilst found to grow through fern cover in other studies, failed to establish in the areas of almost 100% fern cover, despite seeding. Breaking up the root mat without preventing fern regrowth appears to have little value in restoring understory vegetation. Where the fern fronds were regularly clipped, black birch, white pine, and bramble Rubus all germinated and grew. As the treatments were applied to small plots with hand tools, the feasibility of applying these treatments on a practical management scale are uncertain.
Note: If using or referring to this published study, please read and quote the original paper, this can be viewed at: