Factors limiting montane forest regeneration in bracken-dominated habitats in the tropics

Silvia C. Gallegos, Stephan G. Beck, Isabell Hensen, Francisco Saavedra, Denis Lippok, Matthias Schleuning

Research output: Contribution to journalArticlepeer-review

9 Scopus citations


Tropical montane forests are threatened by human-induced fires, resulting in forest patches surrounded by disturbed habitats frequently dominated by bracken fern (Pteridium spp.). In most disturbed habitats, natural forest regeneration is principally hindered by (i) seed limitation, due to a lack of seed dispersal, and (ii) establishment limitation, due to harsh abiotic conditions or competition with grasses and weeds. To our knowledge, the combined effect of seed limitation and establishment limitation on the entire regenerating plant community has never been tested in tropical bracken-dominated habitats. We experimentally assessed the effect of bracken fronds on natural seedling establishment in different distances from a tropical montane forest of the Bolivian Andes. At ten sites, we placed 240 m-long transects ranging from the forest interior to the disturbed habitat and installed paired plots at seven distances along each transect. In one plot of each pair, we applied a repeated vegetation removal treatment to assess the effect of bracken fronds on seedling establishment. We analysed the effect of bracken removal on seedling richness, seedling survival and growth of tree and non-tree species. We additionally compared patterns in seedling density and richness from the forest interior towards the disturbed habitat with patterns in seed density and richness in the seed rain and the soil seed bank. Species richness was higher in the forest than in the disturbed habitat. Tree species corresponded to approx. 25% of the species pool in the disturbed habitat and were only present at low densities. Density and species richness of seeds and tree seedlings consistently decreased from the forest edge towards the disturbed habitat, suggesting seed limitation. Seedlings of tree species had high survival probabilities in the disturbed habitat, especially under bracken fronds, suggesting a lack of establishment limitation. Bracken removal in the disturbed habitat did not affect species richness, but significantly reduced seedling survival and growth of tree and non-tree species. Our results suggest that low seed dispersal hinders natural regeneration of bracken-dominated habitats in the tropics and that bracken fronds do not impede seedling establishment. Seed addition of tree species into the bracken vegetation may therefore be a promising strategy to accelerate forest regeneration, but long-term studies and multispecies experiments are needed for assessing the success of this restoration strategy.

Original languageEnglish
Pages (from-to)168-176
Number of pages9
JournalForest Ecology and Management
StatePublished - 1 Dec 2016
Externally publishedYes

Bibliographical note

Funding Information:
The study was funded by German Science Foundation (DFG; HE 3041/20-1 ) and supported by the research funding programme ‘ LOEWE – Landes-Offensive zur Entwicklung Wissenschaftlich-ökonomischer Exzellenz ’ of Hesse’s Ministry of Higher Education, Research, and the Arts. We thank K. Böhning-Gaese for her support. We are very grateful to A. Fuentes, S. Achá, L. Marconi, P. Pozo, R. de Michell, C. Maldonado, F. Miranda and I. Jimenez for their help on plant identification and to D. Bowler for language polishing in the manuscript. We also thank R. Velasco and all the field and laboratory assistants for their support. Two anonymous reviewers made important comments to improve this manuscript. The authors have no conflict of interest to declare.

Publisher Copyright:
© 2016 Elsevier B.V.


  • Establishment limitation
  • Facilitation
  • Forest restoration
  • Pteridium
  • Seed limitation
  • Seedling establishment


Dive into the research topics of 'Factors limiting montane forest regeneration in bracken-dominated habitats in the tropics'. Together they form a unique fingerprint.

Cite this