Costa Rica: Los Arboles Wet Tropical Forest Restoration


This project originated in 1992 when the Tropical Forestry Initiative purchased land in southwestern Costa Rica, in which four families from the eastern United States purchased abandoned pasture land 10 km from Dominical, Costa Rica to study the potential for restoring mixed stands of wet tropical species to the area. Building on studies that showed promise of fast-growing plantation species for jumpstarting succession in tropical forest restoration, the project sought to measure the success of planting mixed species of native species for forest restoration. The project established a nursery with seeds collected from the vicinity and over five years planted the trees in mixed stands in the pastureland, expanding by 3000 to 5000 trees per year. The range of species utilized gradually grew to include 60 species and nearly 30,000 seedlings.

Quick Facts

Project Location:
Dominical, Puntarenas Province, Savegre, Costa Rica, 9.2524311, -83.85923059999999

Geographic Region:
Latin America

Country or Territory:
Costa Rica

Tropical Forest

Tropical Forest - Moist Broadleaf

Area being restored:
145 hectares

Project Lead:
Tropical Forestry Initiative

Organization Type:
NGO / Nonprofit Organization


Project Stage:

Start Date:

End Date:

Primary Causes of Degradation

Agriculture & Livestock, Deforestation

Degradation Description

The original ecosystem was deforested for timber in the 1950s and then converted to grassland and used as pasture over the next 45 years. While there were remnants of the primary forest around the area, the deforestation widely fragmented the forest.

Reference Ecosystem Description

The area was primary forest until the 1950s when it was cleared for timber. After that time the land has been utilized as pasture for low intensity cattle production. This was true for most of the region, where the conversion to pasture happened into the 1960s and 1970s with fragments of primary forest scattered throughout the landscape. The primary forest in this area consisted of lowland tropical wet forest, which was a very complex tropical forest largely determined by its higher rainfall, lower evapotranspiration and low altitude.

Project Goals

The central goal of the project was to demonstrate that planting and managing native tree species accelerated successional rates toward the ecological restoration of wet tropical forest. Secondarily, the project sought to develop appropriate tree nursery management and planting techniques for use locally which would also show that restoration was an ecologically viable land use project that benefited the local economy. The project also sought to conduct sufficient monitoring of the project to assess the restoration of ecological functions that resulted form the reforestation efforts. The project was started with the intent to exchange technical and socioeconomic information related to reforestation with local landowners and organizations who were involved in forestry management.


The project does not have a monitoring plan.


The intention of the project was to show that restoration of a mixed tropical forest could be done and that it would have viable economic benefits for local landowners, while enhancing the ecological functioning of the restored ecosystems. The Tropical Forestry Initiative is a non-profit organization that was formed by eight people of diverse professional backgrounds who shared a concern for the loss of tropical forests. This project emerged out of the desire of this group to demonstrate that forests can be restored by planting and managing native tree species.

Description of Project Activities:
Beginning in 1993, the project began by first gathering local seed stock of seven different native hardwood species from remnant patches of tropical forest in the area and on local farms, the project then grew those seeds out at the nursery which was established. The seeds were germinated in a seed-bed before being transferred to soil in plastic bags of approximately 10cm x 25cm. After three months about 5000 of these seedlings were planted in the pasturelands at a spacing of about 3m x 3m. In the scrub these plantings were done at approximiately 4m x 4m or more. In the subsequent 5 years from this initial planting the number of plantings were expanded by 3000 to 5000 annually, until a total of 60 different species had been used and amounted to nearly 30,000 seedlings. For the first four years, the seedlings planted in the field required the clearing of competition for about a meter around each tree. Transects of 750m in length were established in which to sample growth performance of at least 24 specimens of each species that were planted originally. Trees for measurement were each marked with aluminum tags. Measurements of height of the first year's plantings has continued on a yearly basis and dbh has been recorded for those species exceeding 10cm after five years. One area that was planted was newly abandoned and a somewhat degraded grass pasture whose aspect was to the northeast. It was planted with seedlings of gallinazo and cedro, among newly established seedlings of mayo Colorado. Another area was planted into a somewhat degraded pasture that sloped gently north. It was abandoned five years' previous and was coming up in a scrub species. After partial removal of this scrub the area was planted with a more complex mixture of seven species"”cenizaro, cocobolo, cristobal, espavel, guanacaste, guapinol, and iguano in the shade of the remnant scrub. As the young trees developed, researchers measured the tree heights with a telescoping meter bar and later with a laser hypsometer. Tree diameters were measured at breast height (dbh). Researchers measured stand density in 0.1-ha plots, using a modification of the strip plot method described by Wenger (1984). This involved five parallel strips of 25 m X 8 m along a 50-m array within a 30-acre (12-ha) block. Researchers calculated wood volumes for Area A using a 0.1-ha plot of the two fastest growing species"”gallinazo and mayo Colorado. For six sample trees of each species, stem profiles were measured using the laser hypsometer combined with a 30-x telescope with a graded reticle and an attached Suunto clinometer. Researchers used the hypsometer to measure the distance to each of five to seven intervals up the stem, and then used the telescope to measure the stem diameter at each interval. From this data, it was possible to determine the average tree profile. For each tree in the plot, the stem volumes were then calculated as the tapered cylinder of the stem defined by the basal area and the height to the first branch. The method is able to measure stem volumes in standing trees. To estimate the wood volume in the remnant primary forest, we counted all the trees that had a dbh greater than 3.9 inches (10 cm) within a circular area of about 0.5 acres (0.2 ha). Since tree profiles in the mature forest were much too complex to measure, researchers made a rough and very conservative estimate of volumes for each tree using the basal area and a simple tapered cylinder as their model. It was not used to estimate the volume of the branches. Instead researchers used tree density and basal area data a 2001 study of the two nearby remnant primary forests. For an index of the species diversity, one researcher surveyed the understory plant species at various sites. In that instance the researcher collected and identified tbe understory species that had appeared in planted areas A and B, and, for comparison, the understory species that could be collected in the two primary forest remnants on the same property. She did not count epiphytic species or lianas. Using data from areas A and B, the project found that the average yearly growth in height of gallinazo and mayo Colorado trees was outstanding"”7.2 ft (2.2 m) and 7.9 ft (2.4 m), respectively. These mean growth rates in height are markedly greater than the growth rates for any of the ten native and exotic tree species that a 1992 study recorded in a study of plantation tree plantings in southern Costa Rica. The fast growth rate of these two species results in a rapid crown closure and, it was assumed, a corresponding abrupt microclimate change.

Ecological Outcomes Achieved

Eliminate existing threats to the ecosystem:
Using data from areas A and B, researchers found that the average yearly growth in height of gallinazo and mayo Colorado trees was outstanding"”7.2 ft (2.2 m) and 7.9 ft (2.4 m), respectively. These mean growth rates in height ate markedly greater than the growth rates for any of the ten native and exotic tree species recorded in a 1992 study of plantation tree plantings in southern Costa Rica. The fast growth rate of these two species results in a rapid crown closure and, it was assumed, a corresponding abrupt microclimate change. Cenizaro, Cristobal, espavel, and guanacaste were intermediate in their height growth rate, each with rates of slightly less than 2.9 ft (0.9 m) per year. Iguano, being a shade-loving species, was slower, growing at about 2.3 ft (0.7 m) per year. The data for cocoholo and guapinol were for a shorter time period, although their growth rates seem to be similar to the four intermediate species. Researchers did not systematically measure cedro after the fourth year because attacks by the Hypsipyla budworm became excessive, and many of the cedro trees became malformed or died. The evidence suggests that the project is making progress toward the goal of establishing a primary forest. This evidence can be seen in terms of the tree density, tree height, tree diameter, basal area and volume, and species diversity. Area A has 880 trees/ha, which is somewhat higher than the average tree density of 750 trees/ha in the local primary forests. A 1997 study reported significantly lower densities for primary forests in the western tropics"”214 to 235 trees/ha. The planted trees have achieved an overall average height of 59.7 ft (18.2 m), neatly two-thirds the height of primary' forest in the neighborhood. Their average dbh of 8 inches (20.5 cm) is about one-third that of the local primary' forests, which is 20.6 inches (52.9 cm). Total basal area of Area A is 31.4 sq m/ha, roughly one-half of the value for primary forests (58.3 m sq/ha). Our estimate of the total wood volume at 32 sq m/ha is about one quarter the estimate for the primary forests, and far below the volumes estimated for Puerto Rico of 145 to 365 sq m/ha. One researcher's 99 understory species indicates aggressive plant succession and increasing bio-complexity under the planted trees. The average count of undergrowth species in two primary forests was somewhat lower at 72 species (not including epiphytes or lianas). A 2004 study also noted that 40 of the understory species in the planted area are common in the local primary forest. These data indicate that the restoration has reached between one quarter and two-thirds of the status of the primary forest, depending on what type of variable is compared. In the instance of the 2004 study, the estimate of progress in the restoration effort seemed to be reflected in the richness of the understory species in common with the primary forest. The range of commonality was between 20-33% overall, which may be a reflection of animal-dispersed characteristic, combined with the great abundance of birds and bats to function as dispersal agents. Wind-dispersed or ejection-dispersed species are much less frequent, but overall the voluntary occurance of man understory species has been surprisingly rapid.

Factors limiting recovery of the ecosystem:
While the diversity of the restoration is far in excess of the monocultural exotic plantations, or the pastureland, it still has a significant way to go toward recovery of the original levels of biodiversity. Some estimates of this process peg the timeframe at between 100 and 800 years. The researchers predict that the relatively rapid of maturation at the outset will gradually slow over time, but after 100 years plantations of mixed tree species may achieve a forest with species composition similar to the primary forest. Another valid concern is the replacement of historic forestry regimes with more sustainable methodology, so as not to replicate the original injury of the 1950s, 60s, and 70s. The move toward inclusive community stewardship, coupled with a broad environmental education enterprise makes this move toward more sustainable methods far likelier.

Socio-Economic & Community Outcomes Achieved

Economic vitality and local livelihoods:
The project was committed to working with local landowners, governmental officials, and schools in an exchange of ideas related to ecological and socioeconomic issues, as well as to promote good will amongst local and foreign-born neighbors. This is an especially important aspect of the work since the Rio Guabo valley and Dominical region have been undergoing a rapid cultural transition due to resort development, improved transportation and communication systems, and access to electrical power. Local landowners and the project worked to establish a forestry association. The goals of the association are centered around promoting reforestation through economic incentives and also to preserve critical habitat. The association's officers and board of directors are all Costa Ricans from the Rio Guabo valley and vicinity. Currently, the association is awaiting legal status, however in the future it will serve as a vehicle for the exchange of technical, economic, cultural and legal information related to forest management.

Long-Term Management

To assess the recovery of species complexity after planting, a long-term monitoring program has been implemented. The centerpiece of this program is a set of permanent plots in aforestation sites at the Los Arbolos field station. The permanent plot design and monitoring program follow methods modified from the Smithsonian Institution’s `Man and the Biosphere Biodiversity Program’ from Dallmier in 1992. Permanent plots (20 m xô°20 m) were established in representative areas of plantings. Each planted seedling as well as naturally colonizing seedlings and coppice or small trees were identified, mapped, and measured (height and dbh). Additionally, canopy coverage was estimated using a canopy densiometer. Seven of these plots were established on replanted pastures while for comparative purposes, forest composition and structure were measured on nearby old-growth forests using point-quarter transect methods.

Sources and Amounts of Funding

The nonprofit Tropical Forestry Initiative, USDA Forest Service,

Other Resources

Tropical Forestry Initiative
P.O. Box 285
Ithaca NY 14851

Primary Contact

Organizational Contact