Content uploaded by Florencia Montagnini
Author content
All content in this area was uploaded by Florencia Montagnini on Oct 10, 2018
Content may be subject to copyright.
Performance of forest plantations in small and medium-sized
farms in the Atlantic lowlands of Costa Rica
Daniel Piotto
a,1
, Florencia Montagnini
b,*
, Luis Ugalde
a
, Markku Kanninen
a
a
Centro Agrono
Âmico Tropical de Investigacio
Ân y Ensen
Äanza (CATIE), 7170 Turrialba, Costa Rica
b
Yale University, School of Forestry and Environmental Studies, 370 Prospect St., New Haven, CT 06511, USA
Received 19 December 2001
Abstract
Exotic tree species predominate in reforestation in tropical regions worldwide. However, some native species are suitable for
plantations, providing a wider variety of products. Adequate government programs of incentives, coupled with good technical
advice to farmers, are needed to stimulate reforestation, especially among small and medium-sized farmers with limited
®nancial resources. This project evaluates growth of native and exotic tree species on plantations in small and medium-sized
farms in the Atlantic humid lowlands of Costa Rica, Central America. A total of 210 pure plantations, ranging in age from 6 to 11
years, were evaluated on 123 farms that had used government incentives for reforestation, and had received technical advice
from local non-government organizations. For each species, seven plantations were chosen at random for study. In each
plantation, plots of 15 trees each were chosen systematically for evaluation of diameter at breast height (dbh), total height,
number of trees per hectare, tree form and spacing. Terminalia amazonia (J.F. Gmel) Exell, Hieronyma alchorneoides Allemao,
and Vochysia guatemalensis Donn. Sm. were the most frequent species found in plantations in the region of study. Gmelina
arborea Roxb. (exotic) and V. guatemalensis (native) had the highest mean annual diameter increment with 2.90 and 2.59 cm,
respectively. Calophyllum brasiliense Cambess. (native) had the lowest mean annual diameter increment (1.48 cm). G. arborea
and V. guatemalensis had the highest mean annual volume increment, followed by Tectona grandis (L.f) Lam. (exotic), and
T. amazonia and Cordia alliodora (R&P) Cham. (both native). Although G. arborea had the greatest mean annual diameter
increment, it had the lowest plantation density and problems with form. V. guatemalensis and T. amazonia, two native species,
were the most promising species for reforestation, due to good growth in volume, good form, and adaptability to a variety of
sites. The poor form of exotic species was related to the low intensity management methods of small farmers. Exotic species had
the highest performance variability between sites, while native species showed relatively high growth homogeneity. Therefore,
native species seem more promising than exotics for general use across varying ecological conditions.
#2002 Elsevier Science B.V. All rights reserved.
Keywords: Costa Rica; Degraded sites; Incentives; Native species; Reforestation
1. Introduction
On a global scale, native species arewidely extracted
from natural forests. However, use of native species for
reforestation is minimal, and exotic tree species pre-
dominate both in industrial as well as in rural devel-
Forest Ecology and Management 175 (2003) 195±204
*
Corresponding author. Tel.: 1-203-436-4221;
fax: 1-203-432-3929.
E-mail address: florencia.montagnini@yale.edu (F. Montagnini).
1
Current address: Rua Afonso Bra
Âs 295, Vila Nova Conceicao,
Sao Paulo, SP, Brazil, CEP 0451-011.
0378-1127/02/$ ± see front matter #2002 Elsevier Science B.V. All rights reserved.
PII: S 0378-1127(02)00127-5
opment plantations (Evans, 1999). In humid tropical
regions of Latin America, forest plantations provide
wood, as well as various environmental services, such
as carbon ®xation, soil stabilization, maintenance of
biodiversity, and restoration of degraded areas. To
ensure success of reforestation programs, farmers must
have access to adequate technical assistance for species
and site selection. In addition, effective programs of
government incentives to stimulate reforestation are
needed, especially for small and medium-sized farmers
with limited ®nancial resources (Segura, 2000).
In Costa Rica over the past 15 years the supply of
wood from natural forests has diminished due to
deforestation combined with the establishment of con-
servation areas (Gonza
Âlez and Fisher, 1994). In
response, more forest plantations have been established
for timber, and a number of these have been established
by small and medium-sized landowners through credit
and government incentives (Watson et al., 1998). Until
1987, one native species (Cordia alliodora (R&P)
Cham.), and three exotics (Gmelina arborea Roxb.,
Pinus spp. and Eucalyptus spp.), made up 94% of the
plantation productionon the Atlantic lowlands of Costa
Rica (Direccio
Ân General Forestal, 1987).
In Costa Rica studies on native species for use in
plantations began in the mid 1980s with research
conducted at La Selva Biological Station of the Orga-
nization for Tropical Studies (OTS) (Butter®eld, 1990;
Espinoza and Butter®eld, 1990). Their research shows
that several native species that had not previously been
used for reforestation had good growth and adapt-
ability to degraded soils.
Between 1987 and 1996, the Costa Rican govern-
ment promoted reforestation through economic incen-
tives, such as the Fund for Forest Development (FDF)
and the Forest Payment Certi®cate (CAF) (Segura
et al., 1996; Watson et al., 1998). Farmers in the
Atlantic lowlands region, in association with local
non-governmental organizations (NGOs): the County
Agricultural Center of Sarapiquõ
Â(CACSA), and the
Foundation for the Development of the Central Vol-
canic Range (FUNDECOR), began establishing forest
plantations. CACSA is an organization made up of
small farmers and supported by the Costa Rican
Ministry of Agriculture and Livestock (MAG), while
FUNDECOR provides technical advice to farmers in
sustainable forest management and reforestation.
When CACSA and FUNDECOR began working on
reforestation projects, they recurred to the OTS for
assistance in species selection. Species were selected
for their promising performance and rapid growth
(Table 1). With the work of CACSA and FUNDECOR,
reforestation in the region passed from a phase of pure
investigation to a phase of implementation (FUNDE-
COR, 2001). By 1999 the projects were responsible
for planting approximately 1500 ha in the Atlantic
lowlands (FUNDECOR, 2001).
However, the investigations at the Biological Sta-
tion La Selva, administered by OTS, only covered the
®rst phases of plantation development (3 and 4 years).
Information about the long-term performance of the
plantations is necessary to guarantee better results and
diminish the investment risk for the rural farmers. The
objective of the present study was to evaluate the
growth of native and exotic tree species in plantations
ranging in age from 6 to 11 years. The plantations
were established by farmers using government incen-
tives, and with the assistance of the local NGOs in the
Atlantic humid lowlands of Costa Rica.
2. Materials and methods
2.1. Site description
The plantations were established between 1990 and
1995 on the Atlantic lowlands of Costa Rica. The study
sites are located at 108120±108470north latitude and
848090±838450west longitude (Fig. 1). Mean annual
temperature is 24 8C. Mean annual precipitation is
3500±5000 mm, and in no month is precipitation less
than 50 mm. Elevation is between 30 and 200 m. The
overall topography is ¯at to undulating terrain. In
general, soils belong to the Ultisol and Inceptisol
orders. There are various limitations of the soil, such
as slow or impeded drainage, and very low to medium
fertility. These limitations restrict land uses to per-
manent crops and reforestation. However, projects
require careful management because of the area's high
susceptibility to erosion (Costa Rica, 1979).
2.2. Characteristics of the farms used in the
present research
This study examines 210 pure plantations on 123
farms, as part of a larger study on farmers' preferences
196 D. Piotto et al. / Forest Ecology and Management 175 (2003) 195±204
for reforestation in the region (Piotto, 2001). Fig. 2
shows total area planted for each species, and the
number of projects established between 1990 and
1995 by the local farmers with assistance from the
local NGOs (CACSA and FUNDECOR).
The majority of the farmers (60%) had less than
10 ha, and only 10% had farms larger than 50 ha. The
plantations had areas between 0.1 and 5.0 ha. Table 2
shows the size and number of farms where the study
areas were located. The majority of the farms were
used for agriculture (bananas, plantains, heart of palm,
maize, cassava, pineapples, black pepper, and a few
ornamentals) (60%), and less for cattle. On 40% of the
farms, the only use of the land was reforestation. The
principal previous land use on reforested lands was
pastures (57%), followed by agriculture (24%) and
young secondary forest (19%). No new areas were
deforested for the establishment of the plantations.
The farms were primarily on undulating terrain (57%),
followed by ¯at terrain (43%). None of the farms were
on steeply inclined terrain.
All the farmers had land titles and received govern-
ment incentives for plantations (22% FDF and 78%
CAF). Incentives were paid in cash and farmers
bought supplies for their plantations individually.
Economic bene®t from harvesting wood was the
Table 1
Species most used by small and medium farmers in the Atlantic humid lowlands of Costa Rica
a
Species Family Natural distribution
Calophyllum brasiliense Cambess Clusiaceae Mexico to N. South America
Vochysia guatemalensis Donn. Sm. Vochysiaceae Mexico to Peru
Hieronyma alchorneoides Allemao Euphorbiaceae Mexico to Brazil
Virola koschnyi Warb. Myristicaceae Central America
Cordia alliodora (Ruiz and Pavo
Ân) Cham. Boraginaceae Mexico to South America
Terminalia amazonia (J. Gmell) Exell. Combretaceae S. Mexico to N. South America
Tectona grandis (L.f.) Lam. Verbenaceae Asia, Malaysia and Indochina
Dipteryx panamensis (Pittier) Record and Mell Fabaceae Nicaragua to Colombia
Gmelina arborea Roxb. Verbenaceae South Asia
a
Jime
Ânez and Poveda (1997), and Carpio (1992).
Fig. 1. Map of Costa Rica showing location of study sites.
D. Piotto et al. / Forest Ecology and Management 175 (2003) 195±204 197
motivation of the majority of farmers (62%) (Piotto,
2001). Nonetheless, some of the farmers reforested for
environmental reasons (38%). These farmers planned
to keep reforested areas as permanent reserves.
2.3. Methods
All plantations in the region were identi®ed and
then a random subset of seven plantations was selected
for study of each species (9 species7 plantations
each 63 plantations total). The number of planta-
tions was decided in order to obtain a 20% sampling
error, which is the highest acceptable error allowed in
forest inventories by the Costa Rican government. The
study used the mean annual increment of diameter at
breast height, dbh variable for Virola koschnyi Warb.
using data already available for this species from
FUNDECOR. This variable was used because the
plantations in this study are of different ages. If basal
area or volume had been used to study plots of
different ages and different spacing between trees,
outcomes would have lacked uniformity.
The following formula was used to calculate the
number of plantations needed to perform this study:
nCV2t2
0:05
SE2
where nis the number of plantations, CV the coef®-
cient of variation (%), t
0.05
the values from the Stu-
dent's t-test at 5% probability, and SE the sampling
error (%).
Because of the lack of information for other species,
this study used a calculated size of the samples for
V. koschnyi and applied this value to all species. How-
ever, differences between species were expected, and
each species had its own sampling error (Table 3). The
table shows that only in the G. arborea and Dipteryx
panamensis (Pittier) plantations was sampling error
above 20%. Because this study was meant to be an
exploratory study comparing growth of different spe-
cies, the study area was not enlarged to address this
problem. Species like Calophyllum brasiliense Cam-
bess, which has nearly the same number of plantations
as T. grandis in the region, had a sampling error of
10%. The species with the largest number of planta-
tions in the area, Terminalia amazonia (J. Gmell)
Exell, had the lowest sampling error (Table 3).
For this project, a multi-series sampling design was
used for each species. The primary units of study were
the plantations, chosen at random. The secondary
Fig. 2. Number of plantations and total planted area for the species most used by farmers with the assistance of local NGOs in the Atlantic
humid lowlands of Costa Rica (plantations established between 1990 and 1995). CA, Cordia alliodora; CB, Calophyllum brasiliense;DP,
Dipteryx panamensis; GA, Gmelina arborea; HA, Hieronyma alchorneoides;TA,Terminalia amazonia; TG, Tectona grandis;VG,Vochysia
guatemalensis; VK, Virola koschnyi.
Table 2
Size and number of farms where tree measurements were taken in
the Atlantic humid lowlands of Costa Rica
Farm size (ha) Number of
farms sampled
Mean area of
farm sampled (ha)
10 25 0.98
10±50 13 1.10
>50 5 2.68
198 D. Piotto et al. / Forest Ecology and Management 175 (2003) 195±204
units of study were temporary plots of 15 trees each.
The 15-tree plots were chosen systematically.
Plots were designed by number of trees because
trees on the plantations showed homogenous spacing,
but with different arrangements (3 m 3m;3:5m
3:5m and 4m4 m). This design guaranteed a uni-
form number of observations per plot, independent of
initial tree spacing. The use of 15-tree plots is sug-
gested by Wright (1964), who compared performance
of 15-tree plots with 200-tree plantation plots. Wright
demonstrated that both plot sizes gave similarly valid
estimations of tree growth relative to height and
diameter.
Systematic plot placement was used in this inves-
tigation to diminish cost and time related to the
investigation. In the majority of the cases, the sys-
tematic approach produces more precise results than
random sampling (Kleinn and Pe
Ârez, 2000). A random
procedure was used to place the ®rst corner of the ®rst
tree-plot, and the following tree-plot was located 50 m
from the initial tree plot. Sampling intensity was
four plots per hectare, representing 5.4% of the area
where plantations were at 3 m 3 m planting dis-
tance, and 9.6% when plantations were at 4 m 4m
spacing.
2.4. Data analysis
In each plot, dbh, total height, number of trees per
ha, tree form, and spacing, were evaluated. Trees
were placed in one of the three categories of form
(straight stems, slightly sinuous or very sinuous
stems). Bifurcation in the stem at height less than
4 m was also recorded. The MIRA-SILV software
was used to analyze the ®eld data (Ugalde, 2000).
Tree survival, basal area and volume were calculated
for each plot. Calculation of volume was done using a
0.5 volume factor (Newbould, 1967). Analysis of
variance was used to compare mean annual incre-
ments in dbh, height and volume, and survival
between species. Means were compared using the
Turkey H SD test.
3. Results
3.1. Plantation spacing
Plantations with spacing of 3 m 3m;3:5m
3:5m and 4m4 m were found for each species.
The most common spacing scheme was 3 m 3 m for
all species. For G. arborea,3m3 m was the only
spacing scheme used. D. panamensis was the only
species with plantations in all three spacing schemes,
including 3:5m3:5m.
3.2. Tree survival and number of trees per hectare
Survival of species in this study re¯ected the per-
centage of trees remaining in plantations after replant-
ing, thinning and natural mortality. The species with
the highest survival was C. brasiliense, with 87.1%
of the trees initially planted still living. The species
with the second highest survival was V. guatemalensis
with 75.1%. G. arborea had the lowest survival with
59.5%. This is a signi®cant difference P<0:01when
compared with C. brasiliense and V. guatemalensis
(Tab le 3).
The number of trees per hectare, which represents
the density of the stands, had low variation without
signi®cant difference between species P<0:01.
The exception to this observation was C.brasiliense
which had the highest density.
3.3. Tree growth and productivity
Signi®cant differences were found P<0:01in
mean annual increments in dbh and mean annual
Table 3
Survival and number of trees per hectare and sampling error on
small and medium farms in Atlantic humid lowlands of Costa Rica
a
Species Survival
(%)
Number of
trees/ha
Sampling
error (%)
C. brasiliense 87.1 a 840 a 9.8
V. guatemalensis 75.1 a 647 b 10.0
H. alchorneoides 72.4 ab 634 b 10.3
V. koschnyi 72.4 ab 681 ab 12.9
C. alliodora 71.9 ab 693 ab 15.7
T. amazonia 69.0 ab 597 b 16.0
T. grandis 68.0 ab 689 ab 17.0
D. panamensis 66.6 ab 654 b 22.2
G. arborea 59.5 b 653 b 26.8
a
For each variable, differences among means in the same
column followed by the same letter are not statistically signi®cant
P<0:01.
D. Piotto et al. / Forest Ecology and Management 175 (2003) 195±204 199
increments in height between species (Table 4). G.
arborea and V. guatemalensis were the species with
greatest mean annual dbh increments, with 2.90 and
2.59 cm, respectively. Differences among these
growth rates were highly signi®cant P<0:01when
compared with other species in the study. T. amazonia
and T. grandis were the exceptions, with similar dbh
growth of 2.36 cm per year. C. brasiliense had the
lowest mean annual dbh increment (1.48 cm). This
was signi®cantly inferior P<0:01to other species,
with the exceptions of Hieronyma alchorneoides
Allemao and D. panamensis (with 1.85 and 1.73 cm,
respectively).
G. arborea,T. amazonia and D. panamensis had
the largest mean annual increment in height with
2.24, 1.97 and 1.94 m, respectively. G. arborea and
V. guatemalensis had the highest mean annual incre-
ment in volume, followed by T. grandis,T. amazonia
and C.alliodora.C. brasiliense and D. panamensis
had the lowest mean annual increment in volume.
3.4. Tree form
V. koschnyi and V. guatemalensis had the highest
proportion of straight stems, and the lowest proportion
of sinuous or bifurcated trees. H. alchorneoides and
T. grandis had the highest percentages of bifurcated
stems with 14.7 and 8.5%, respectively. G. arborea
also had inferior stem form when compared with
native species, despite being the species with the
highest growth (Table 4).
4. Discussion
4.1. Promising species for reforestation: productivity
and economic and environmental values
Use of exotic species predominates in tropical
timber and rural development plantations worldwide
(Evans, 1999). However, in the Atlantic lowlands of
Costa Rica, 88% of plantations established between
1990 and 1995 used native species. The high use of
native species in the region is largely due to the results
of tree planting projects at La Selva Biological Station,
as well as to the technical assistance regarding species
selection provided to farmers by the local NGOs.
Projects at La Selva demonstrated that some native
species, which had not previously been used for
reforestation, were in fact appropriate for silviculture
(Butter®eld, 1990; Espinoza and Butter®eld, 1990). In
addition, the government subsidies provided at the
time were effective in promoting reforestation with
native species by small and medium farmers through-
out the region.
The results of this investigation show that the
majority of plantations in the area had grown well,
and some advantages of using native species over
exotics were apparent. Although an exotic species
(G. arborea) had one of the greatest annual growth
in dbh and height, it had the lowest plantation density
and problems with tree form. This species has been
studied extensively in the neotropics, and similar
results have previously been reported in Costa Rica
Table 4
Mean annual increment in diameter (MAIDBH), mean annual increment in height (MAIH), mean annual increment of volume (MAIV) and
tree form of the nine species most used by small and medium farmers in the Atlantic lowlands of Costa Rica
a
Species MAIDBH
(cm per year)
MAIH
(m per year)
MAIV
(m
3
/ha per year)
Straight
stem (%)
Slightly
sinuous (%)
Very
sinuous (%)
Forked
trees (%)
G. arborea 2.90 (0.49) a 2.24 (0.42) a 31.49 (14.3) a 86.2 12.2 1.6 0.8
V. guatemalensis 2.59 (0.29) a 1.90 (0.26) b 23.51 (13.0) ab 96.8 2.6 0.6 1.3
T. amazonia 2.36 (0.25) ab 1.97 (0.31) ab 17.66 (7.16) bcd 96.6 3.4 0.0 7.6
T. grandis 2.36 (0.57) ab 1.88 (0.42) b 21.39 (12.7) b 78.9 17.6 3.5 8.5
C. alliodora 2.09 (0.38) bc 1.68 (0.32) bc 15.67 (6.36) bcde 86.0 12.0 2.0 4.0
V. koschnyi 1.95 (0.34) bc 1.41 (0.37) c 11.47 (7.67) cdef 100.0 0.0 0.0 0.0
H. alchorneoides 1.85 (0.26) bc 1.71 (0.27) bc 10.21 (6.13) def 84.0 15.3 0.7 14.7
D. panamensis 1.73 (0.50) c 1.94 (0.47) ab 7.90 (4.39) ef 89.2 10.1 0.7 5.8
C. brasiliense 1.48 (0.16) c 1.29 (0.19) c 5.95 (3.13) f 95.1 4.9 0.0 4.4
a
Differences among means for each variable are statistically signi®cant when the standard deviation is followed by different letters
P<0:01.
200 D. Piotto et al. / Forest Ecology and Management 175 (2003) 195±204
(Centro Agrono
Âmico Tropical de Investigacio
Ân and
Ensen
Äanza, 1991; Gonza
Âlez and Fisher, 1994).
Vochysia guatemalensis and T. amazonia were the
most promising native species for reforestation, with
annual growth comparable to the most commonly used
exotic species (Martõ
Ânez, 1981). Similar results for
younger trees of these species have been reported in
other studies conducted in the same region (Butter®eld
and Espinoza, 1995; Gonza
Âlez and Fisher, 1994;
Espinoza and Butter®eld, 1990). The results of this
investigation are therefore, in accordance with pre-
vious projects, demonstrating the viability of using
native species for timber-producing reforestation in
the region. Although at present the total rotation times
for these species are not known, expected volumes at
harvest are estimated as 170 m
3
/ha for T. amazonia
after 25 years, and 250 m
3
/ha for V. guatemalensis
after 20 years (Montagnini, 2001). These yields make
them attractive alternatives for farmers. Fuelwood
from thinnings represents an additional source of
income for the farmers (Montagnini and Mendelsohn,
1997).
The four species with best growth were two native
species (V. guatemalensis and T. amazonia) and two
exotics (G. arborea and T. grandis). Nonetheless,
when comparing tree form, native species were super-
ior. The poor form of exotic species could be related to
the low intensity of management applications by small
farmers (Haggar et al., 1998). Normally, these exotic
species require intensive management to guarantee
productivity and high wood quality.
In the present research, for each exotic species,
there is a native species that is comparable in wood
quality and growth. G. arborea (exotic) and V. gua-
temalensis (native) produce light wood with speci®c
weight of approximately 0.34 and 0.35 g/cm
3
, respec-
tively. T. grandis (exotic) and T. amazonia (native)
produce heavier wood, with mean speci®c weights of
0.61 and 0.68 g/cm
3
, respectively (Carpio, 1992). Use
of native species, which are better adapted to local
conditions, instead of exotics is one way to maintain
wood production and increase the conservation value of
plantations (Keenan et al., 1999; Parrotta et al., 1997).
In addition, native species are better suited to low-input
forestry practices (Haggar et al., 1998), common on
small and medium-sized farms in tropical regions.
D. panamensis and C. brasiliense had the lowest
growth in dbh and height, and could be classi®ed as
slow-growing species. These species need long rota-
tion periods to produce saw timber. Estimated rotation
times are 35±40 years, with estimated volumes at
harvest of 200 m
3
/ha for C. brasiliense and 300 m
3
/
ha for D. panamensis (Montagnini, 2001). C. brasi-
liense was the slowest growing species, with growth
rates slightly lower than those found by Butter®eld and
Espinoza (1995) and by Montagnini et al. (1995) in 4-
year-old pure plantations. Nonetheless, many small
farmers have chosen to plant these species because of
the high priced wood and their good form. The high
value of the wood compensates for the low annual
increments in volume. These species also have a high
social and environmental value. D. panamensis seeds
are an important food source for fauna, including the
green macaw (Ara ambigua) (an endangered species)
and some rodents like Sciurus granantensis,Dasy-
procta punctata,Agouti paca, and Proechimys semi-
spinosus.D. panamensis also contributes to the
establishment of other vegetation because of its inter-
actions with several seed dispersers and pollinators
(Bonaccorso et al., 1980; Terborgh and Wright, 1994).
H. alchorneoides,C. alliodora and V. koschnyi are
other native species that had relatively good growth,
although lower than the exotic species (G. arborea and
T. grandis) and than the other more productive native
species of this research, such as V. guatemalensis and
T. amazonia.H. alchorneoides has a relatively high
percentage of bifurcated trees (14.7%) and some
problems with tree form. However, estimated revenues
from harvesting H. alchorneoides are high due to
relatively high market prices, making its planting an
attractive alternative (Montagnini and Mendelsohn,
1997). This species requires intensive management
during initial phases of a plantation development, such
as pruning, to guarantee higher quality stems for ®nal
harvest. Nonetheless, for small farmers interested in
high value timber, these species could be used to
obtain a greater variety of products to respond to
market uncertainty (Montagnini et al., 1997).
4.2. Effects of site and silvicultural practices
on species performance
The sites where the plantations were established are
marginal lands for traditional agriculture because of
their poor and degraded soils. In these zones, as in
many regions of the humid tropics worldwide, forestry
D. Piotto et al. / Forest Ecology and Management 175 (2003) 195±204 201
activities are relatively new, and are generally prac-
ticed by small and medium farmers (5±50 ha farms).
These farmers generally use low inputs, give little
preparation to the sites, and manage the plantations
with low intensity. Varying site conditions and low
intensity management can create a high variability in
productivity of the same species on different planta-
tions. For higher growth homogeneity across planta-
tions, foresters should use more stable genotypes
(Butter®eld, 1996). That is, genotypes that maintain
a certain level of productivity across varying condi-
tions should be used. Using dbh, mean annual height
increment, and variance between plantations (S
2
)
(Fig. 3), this study provides insight as to which species
have higher performance homogeneity.
Exotic species had the highest performance varia-
bility across sites and under different management
plans. Exotic species had relatively higher growth in
height and dbh, but also had high variability between
sites (Fig. 3). These ®ndings indicate that these exotic
species could be considered ``specialists'', i.e., site
variability and management have a substantial impact
on growth and development of these species. None-
theless, if the variables affecting growth could be
identi®ed, the exotic species could be planted in
adequate sites and managed in a way to maximize
productivity (Butter®eld, 1996).
On the other hand, native species had relatively high
growth homogeneity across the entire study area, with
the exception of D. panamensis. Native species seem
more promising than exotics for general use across
varying ecological conditions. For example, V. gua-
temalensis and T. amazonia, in addition to having
similar growth rates to the exotic species and good
expected volume yields at harvest, displayed high
plasticity in relation to site conditions. These two
species seem to be the best alternative for local
plantations, as well as for plantations in other regions
with similar ecological conditions.
The rest of the native species, which displayed
relatively lower growth rates but high plasticity, could
be used on less favorable sites. These species should
be planted on soils that are compacted from previous
land use, have poor drainage, low fertility, and in areas
where there is limited management.
4.3. Farmer objectives and needs for the
promotion of reforestation
Previous studies indicate that many farmers
throughout the tropics perceive reforestation as a
way to make marginal farm areas productive, both
environmentally and economically (Dõ
Âaz, 1995; Kor-
honen, 2000). The primary objective of plantations in
this study was to bring economic bene®ts to farmers
and increase forest products for farm use. Environ-
mental objectives have prevailed in importance in
previous studies in Costa Rica (Martõ
Ânez et al.,
1994; Current and Scherr, 1995; Dõ
Âaz, 1995; Schelhas
et al., 1997; Thacher et al., 1997). A farmer's objective
in establishing a plantation is associated to his/her
Fig. 3. Variance (S
2
) versus mean annual increment in dbh
(MAIDBH) and mean annual increment in height (MAIH) for the
nine species most used by farmers in the Atlantic humid lowlands
of Costa Rica. CA, Cordia alliodora; CB, Calophyllum brasiliense;
DP, Dipteryx panamensis; GA, Gmelina arborea; HA, Hieronyma
alchorneoides;TA,Terminalia amazonia; TG, Tectona grandis;
VG, Vochysia guatemalensis; VK, Virola koschnyi.
202 D. Piotto et al. / Forest Ecology and Management 175 (2003) 195±204
economic situation and level of education. Poorer
farmers plant trees for economic bene®t and domestic
needs. Current and Scherr (1995) found that economic
bene®t is the primary reason for the establishment
of plantations in Central America, where high levels
of poverty predominate, especially in rural areas
(Kaimowitz, 1996).
Most farmers in the region believe that the govern-
ment established forest policy for long-term planning
(Piotto, 2001). They believe that the government must
guarantee ®nancial support to ensure that plantations
complete a productive rotation. The government must
also simplify bureaucratic procedures for distribution
of incentives. Many studies have demonstrated that in
most developing countries, projects developed and
maintained with incentives are not sustainable after
®nancial support ends (Roche, 1997).
The forestry sector needs greater emphasis on social
implications of reforestation, especially because
incentives tend not to bene®t the poorest farmers
(Watson et al., 1998). Small scale reforestation pro-
jects are apparently more expensive, but small projects
are a better way to distribute bene®ts. Without ®nan-
cial support, forest resources on farms will continue to
be managed unsustainably.
Farmers in the Atlantic lowlands of Costa Rica
are willing to continue reforestation (Piotto, 2001).
Forestry is commonly perceived to provide many
goods and services. However, farms have low income,
limiting the establishment of plantations due to long
rotations and high risk (Segura, 2000). Therefore,
programs and projects to support reforestation on
small farms should be improved to satisfy social,
economic and environmental goals.
5. Conclusions
Widespread use of native species for reforestation
programs in the Atlantic lowlands of Costa Rica was,
as the results of this study show, a good decision.
Native species had better or equal performance than
exotic species, and also had greater adaptability to a
variety of site conditions than exotic species. Use of
native species helped create greater production diver-
si®cation on the plantations. Also, the use of medium
and high quality native species in reforestation pro-
jects can alleviate logging pressure on natural forests.
The majority of native species used demonstrated
high potential productivity for commercial planta-
tions. The native species, particularly V. guatemalensis
and T. amazonia, showed rapid growth, high survival
rates, and plasticity to respond to different site con-
ditions. Therefore, these species could also be used
successfully in reforestation in other regions with
similar ecological conditions. Programs for genetic
improvement should be considered to increase pro-
ductivity and reduce rotation cycle length. Further
work to improve native species for commercial use
will make them more attractive to small farmers, and
reduce the need for government incentives.
Acknowledgements
This project was funded by FINNIDA. Daniela
Cusack translated the manuscript into English. M.
Ashton and M. Wishnie gave useful comments on
this article. The authors thank CACSA and FUNDE-
COR for their generous assistance.
References
Bonaccorso, F.J., Glanz, W.E., Sandford, C.M., 1980. Feeding
assemblages of mammals at fruiting Dipteryx panamensis
(Papilionaceae) trees in Panama: seed predation, dispersal, and
parasitism. Rev. Biol. Trop. 28, 61±72.
Butter®eld, R., 1990. Native species for reforestation and land
restoration: a case study from Costa Rica. In: Proceedings of
the 14th IUFRO World Congress, Vol. 2, Montreal, Canada,
pp. 3±14.
Butter®eld, R., 1996. Early species selection for tropical reforesta-
tion: a consideration of stability. For. Ecol. Manage. 81, 161±
168.
Butter®eld, R., Espinoza, M., 1995. Screening trial of 14 tropical
hardwoods with an emphasis on species native to Costa Rica:
fourth year results. New For. 9, 135±145.
Carpio, I.M., 1992. Maderas de Costa Rica: 150 especies forestales.
Universidad de Costa Rica, San Jose
Â, Costa Rica, p. 338.
Centro Agrono
Âmico Tropical de Investigacio
Ân and Ensen
Äanza,
1991. Melina (Gmelina arborea): especie de a
Ârbol de uso
mu
Âltiple en Ame
Ârica Central. Ser. Te
Âc., Coleccio
Ân Guõ
Âas
Silviculturales No. 10. CATIE, Turrialba, Costa Rica, p. 62.
Costa Rica, 1979. Manual descriptivo de los criterios, clases and
subclases del mapa ``Capacidad de uso de los suelos de Costa
Rica''. O®cina de Plani®cacio
Ân Sectorial Agropecuario, San
Jose
Â, Costa Rica, p. 100.
Current, D., Scherr, S., 1995. Farmer costs and bene®ts from
agroforestry and farm forestry projects in Central America and
Caribbean: implications for policy. Agrofor. Syst. 30, 87±103.
D. Piotto et al. / Forest Ecology and Management 175 (2003) 195±204 203
Dõ
Âaz, Y.V., 1995. Socioeconomõ
Âa y silvicultura del establecimiento
de plantaciones forestales en ®ncas pequen
Äas del Canto
Ânde
Pe
Ârez Zeledo
Ân, Costa Rica. Tesis M.Sc. CATIE, Turrialba,
Costa Rica, p. 94.
Direccio
Ân General Forestal, 1987. Boletõ
Ân estadõ
Âstico No. 2.
Direccio
Ân General Forestal, San Jose
Â, Costa Rica, p. 5.
Espinoza, M., Butter®eld, R., 1990. Adaptabilidad de 13 especies
nativas bajo condiciones de plantacio
Ân en las tierras bajas
hu
Âmedas de Costa Rica. En: Salazar, R. (Ed.), Manejo y
Aprovechamiento de Plantaciones Forestales con Especies de
Uso Mu
Âltiple. Actas Reunio
Ân IUFRO, Guatemala, Abril de
1989. CATIE, Turrialba, Costa Rica, pp. 159±172.
Evans, J., 1999. Planted forests of the wet and dry tropics: their
variety, nature, and signi®cance. New For. 17, 25±36.
FUNDECOR, 2001. Programas y su impacto: reforestacio
Ân.
Consulted 28 May 2001. Available in http://www.fundecor.-
or.cr/logros/programas/reforestacion.shtml.
Gonza
Âlez, E., Fisher, R., 1994. Growth of native species planted on
abandoned pastureland in Costa Rica. For. Ecol. Manage. 70,
159±167.
Haggar, J.P., Briscoe, C.B., Butter®eld, R.P., 1998. Native species:
a resource for the diversi®cation of forestry production in the
lowland humid tropics. For. Ecol. Manage. 106, 195±203.
Jime
Ânez, Q., Poveda, L.J., 1997. Lista actualizada de los a
Ârboles
maderables de Costa Rica. Aportes al Desarrollo Sostenible.
Universidad Nacional, Heredia, Costa Rica, p. 36.
Kaimowitz, D., 1996. Livestock and deforestation. Central America
in the 1980s and 1990s: a policy perspective. CIFOR Special
Publication. Center for International Forestry Research, Bogor,
Indonesia.
Keenan, R.J., Lamb, D., Parrotta, J., Kikkawa, J., 1999. Ecosystem
management in tropical timber plantations: satisfying econom-
ic, conservation, and social objectives. J. Sustain. For. 9, 117±
134.
Kleinn, C., Pe
Ârez, J., 2000. Consideraciones metodolo
Âgicas en la
experimentacio
Ân cientõ
®ca agrõ
Âcola. Sub-Unidad de Estadõ
Âstica.
CATIE, Turrialba, Costa Rica, p. 28.
Korhonen, K., 2000. The silvicultural state of planted forests in
southern Costa Rica as affected by farmers' motivation for
reforestation: evaluation of forest incentive programs. Thesis.
University of Helsinki, Finland, p. 71.
Martõ
Ânez, H.H., 1981. Evaluacio
Ân de ensayos de especies forestales
en Costa Rica. Tesis. CATIE, Turrialba, Costa Rica, p. 172.
Martõ
Ânez, H.A., Sage, L.F., Borge, C., Picado, W., 1994.
Evaluacio
Ânte
Ânica externa del Programa de Desarrollo Forestal.
DGF-DECAFOR. Secretarõ
ÂaTe
Âcnica de Apoyo. Fondo de
Desarrollo Forestal, San Jose
Â, Costa Rica, p. 120.
Montagnini, F., 2001. Plantaciones forestales con especies nativas:
una alternativa para la recuperacio
Ândea
Âreas degradadas.
Biocenosis (Costa Rica) 15 (1±2), 72±78.
Montagnini, F., Mendelsohn, R., 1997. Managing forest fallows:
improving the economics of swidden agriculture. Ambio 26,
118±123.
Montagnini, F., Gonza
Âlez, E., Rheingans, R., Porras, C., 1995.
Mixed and pure forest plantations in the humid neotropics: a
comparison of early growth, pest damage and establishment
costs. Commonw. For. Rev. 74, 306±314.
Montagnini, F., Sancho, F., Gonza
Âlez, E., Porras, C., Moulaert, A.,
Mo
Ânaco, A., 1997. Plantaciones forestales puras y mixtas con
especies nativas para la reforestacio
Ân de terrenos degradados en
Costa Rica: estudio comparativo del crecimiento, dan
Äo por
plagas, regeneracio
Ân natural y costos de establecimiento.
Biocenosis (Costa Rica) 12, 25±34.
Newbould, P.J., 1967. Methods for Estimating the Primary
Production of Forest. IBP Handbook 2. Blackwell Scienti®c
Publications, Oxford, UK.
Parrotta, J.A., Turnbull, J.W., Jones, N., 1997. Catalyzing native
forest regeneration on degraded tropical lands. For. Ecol.
Manage. 99, 1±7.
Piotto, D., 2001. Plantaciones forestales en Costa Rica y
Nicaragua: Comportamiento de las especies y preferencias de
los productores. Tesis M.Sc. CATIE, Turrialba, Costa Rica.
Roche, L., 1997. Of®cial development aid policies and sustainable
utilization of forest resources in developing countries. Com-
monw. For. Rev. 76, 91±97.
Schelhas, J., Jantzi, T., Kleppner, C., O'Connor, K., Thatcher, T.,
1997. Costa Rica: meeting farmers needs through forest
stewardship. J. For. 95, 33±38.
Segura, O., 2000. Sustainable systems of innovation: the forest
sector in Central America. SUDESCA Research Papers, 24.
Department of Business Studies, Aalgorg, Denmark.
Segura, O., Miranda, M., Mejõ
Âas, R., 1996. Plani®cacio
Âne
inversio
Ân en el sector forestal en Costa Rica. Consejo
Centroamericano de Bosques y A
Âreas Protegidas, Heredia,
Costa Rica, p. 32.
Terborgh, J., Wright, S.J., 1994. Effects of mammalian herbivores
on plant recruitment in two neotropical forests. Ecology 75,
1829±1833.
Thacher, T., Lee, D.R., Schelhas, J.W., 1997. Farmer participation
in reforestation incentive programs in Costa Rica. Agrofor.
Syst. 35, 269±289.
Ugalde, L., 2000. El sistema MIRA, Componente de Silvicultura.
Manual del usuario. CATIE, Turrialba, Costa Rica, p. 82.
Watson, V., Cervantes, S., Castro, C., Mora, L., Solõ
Âs, M., Porras,
I.T., Cornejo, B., 1998. Abriendo espacio para una mejor
actividad forestal. Estudio de Costa Rica. Centro Cientõ
®co
Tropical e Institute for Environment and Development, San
Jose
Â, Costa Rica, p. 114.
Wright, J.W., 1964. Mejoramiento gene
Âtico de los a
Ârboles
forestales. FAO, Rome, p. 437.
204 D. Piotto et al. / Forest Ecology and Management 175 (2003) 195±204
