Natural Regeneration

A Controlled Experiment Shows Similar Early Pathways of Tree Recruitment in Natural Regeneration and Tree Planting Treatments

Gilman, A.C., Letcher, S. G., Fincher, R.M., Perez, A.I., Madell, T.W., Finkelstein, A.L. and Corrales-Araya, F. 2016. Recovery of floristic diversity and basal area in natural forest regeneration and planted plots in a Costa Rican wet forest. Biotropica, 48: 798–808


Download PDF

More than 5 million hectares of tropical forest continue to be lost every year, despite recent declines in t
ropical deforestation rates. Net forest loss would be considerably larger were it not for global reforestation efforts that permit the growth of new forest habitat on unused agricultural land or the development of tree plantations in extensive areas. As governments, NGO´s, and private landowners have limited funds to invest in forest restoration projects, research is needed to address how to best restore rainforest on unused farmland for the lowest cost. Surprisingly, few studies provide a satisfying answer to this question.

Traditionally, forestry-oriented investigations focus on how much timber can be grown on an area of land, and what value that timber will have at the time of harvest to ensure a viable economic return. In contrast, ecological research often focuses on the diversity and types of species that thrive in different age forests or in regrowth forests with different histories of human impacts. Very few studies experimentally compare tree species diversity across different restoration approaches, while controlling for land-use history, landscape surroundings, and regional climate conditions.

One key technique of forest regrowth is natural regeneration, also known as a passive restoration, where an area is fenced off to keep hungry cattle out and let the forest regrow at its own pace with trees that colonize on their own or with help from animal dispersal agents. The second technique is actually planting trees, known as active restoration, where cattle is fenced out, and tree seedlings are laboriously planted throughout the whole area or in dense clumps. Tree plantations create forest cover faster than natural regeneration, but come at a high cost of labor and materials where young seedling trees have to be purchased, transported to the site, planted, and maintained after planting to prevent aggressive pasture grass and vines growing over them, for at least the first 3 or 4 years.  The decision whether to plant trees also raises the question of how many species to plant: just one species across the entire area, or different mixtures of species that have vary in their growth rates, resources demand, and interactions with other species.

This study reports on the first 5 years of results from an experiment we conducted in the wet forest region of NE Costa Rica, established as a long-term study.  We planted our active restoration plots with native species found locally, and used mixtures of 3, 6, 9 and 12 species.  The passive restoration plots, that we call control plots in the study, were randomly planted among the others, with 1-m tall pvc tubes instead of tree seedlings, to allow the same maintenance, recruitment surveys, and bird perches in every plot. The experimental and control plots were established in former cattle pasture and shared the same historical use and soil conditions.  Also, all our plots were close to secondary forest areas over 15 years old and to old-growth forest areas, which we believe had never been cleared. With excellent seed sources nearby, we investigated what new plant species were arriving in both planted and unplanted plots.

After 5 years, the planted plots accumulated greater tree basal area than in natural regeneration plots. This was an expected result from actively planting trees instead of pvc tubes. But when we considered only the new stems that recruited into our plots on their own, we found no difference in the basal area of woody recruits in the planted vs. the control plots. Within the planted plots, the basal area of woody recruits was not affected by the number of species that were mixed together. From this we concluded that if carbon storage or timber is your end goal of your restoration project, active restoration by planting trees provides better short term results compared with natural regeneration. But this finding is only five years into a long-term process and may not hold over time.

Tree species slowly accumulated in the plots over the five years, with similar species recruiting into the natural regeneration and planted plots. This result can be explained by the nearby areas of older forest that provided numerous and diverse wind and animal dispersed seeds into all our plots. At the 5-year point, all our plots seem to be following a similar trajectory, with grass and fern floor cover declining as canopies begin to close and reduce understory light levels. However, 5 years in the life of a recovering forest is a very short window, and we are looking forward to seeing if this trend is maintained over the next 20 years or so, or whether planting more species will, in the long-term, provide more habitat opportunities for incoming species, and potentially speed up forest recovery and the accumulation of new biodiversity over time.


The PARTNERS connection
I met my coauthors during my PhD field research in Costa Rica while we were all living at the La Selva Biological Station. We also met Robin Chazdon at La Selva and in that melting pot our project ideas and long-term collaborations were developed. Robin was aware of our restoration project from its infancy, and after several PARTNERS workshops resulted in the idea of a Special Issue of Biotropica on the topic of natural regeneration in landscape restoration, she kindly invited us to submit a paper. Like our own children, our plots continue to grow and we continue to measure their achievements, hopefully, for many more years to come.