GuANACASTE NATIONAL PARK:

TROPICAL ECOLOGICAL AND CULTURAL RESTORATION

DANIEL H. JANZEN
 GUANACASTE NATIONAL PARK:
TROPICAL ECOLOGICAL AND CULTURAL RESTORATION

Publishedunder the auspiciesof

Servicio de Parques Nacionales (Costa Rica). Fundaci6n de Parques Nacio-
nales (Costa Rica). Fundacion Neotropica (Costa Rica). Universidad Estatal a
Distancia (Costa Rica). Nature ConservancyInternational Program: Costa
Rica (U.S.A.). Tinker Foundation (U.S.A.).
I
 Special Costa Rica edition by Editorial
Universidad Estatal a Distancia,
(EUNED), Sanjose, 1986.

All rights reserved. No partof this book

may be reproduced in any form or
by any means without permission in writing
from the publisher.

639.959.728.6
J35g Janzen,DanidH
Guanacascc National Park: tropical ecological and cultural pasto-
ration / Daniel H . Janzcn .•• 1. ed.·· San Jose, C.R. : EUNED-FPN -
PEA, 1986.
104 p.; mapas; 27 cm.

Incluye apcndicc al final de la obra

ISBN 9977-64-316-4

1. Parques nacionales · Costa Rica. 2. UNED - Costa Rica . I. Tltu·

lo.

Printed in Costa Rica by

Editorial EUNED .
 7

COVER

CENTER. Eastern and southern half of Guanacaste National

Pad (GNP), as viewed from about 1200 meters elevation over the
Pacific Ocean. The lower two-thirds of the photograph is primarily
Santa Rosa National Park. This Park is the primary source of the
dry forest habitats and organisms that will reoccupy much of the
land to be incorporated in GNP.
In the foreground lies Playa Naranjo (Naranjo Beach). In the
lower left is a southern comer of the severely deforested Santa
Elena Peninsula, which extends to the left and rear of this photo-
graph. Moving inland,sthere are two large seasonal river drainages
(center-left, Rio Nisperal; center, Rio Calera). The Santa Rosa
plateau is covered with a mosaic of various ages of deciduous forest
and ab~ndoned jaragua pasture (yellow). The escarpment running
to the left from the center of the photograph is closely paralleled
by the Park entrance road. Further inland lie ranches outside of
Santa Rosa. The upper part of the plateau was previously covered
with oak forest and joins with the bases of Volcan Oros1 (left) and
Volcan Cacao (right). Old but active pastures are visible as yellow
patches cut out of the evergreen pristine forest on the sides of the
volcanos. The clouds are the moisture that generates the headwaters
of the Rfo Tempisque, which flows out through the upper right -hand
corner of the photograph.
UPPER LEFT. Ten-day-old seedlin~ of the guanacaste tree
(Enterolobium cyclocarpum) that are germinating from seeds occur-
ring naturally in cattle dung. Cattle eat the large fruits of this large
tree and are major dispersal agents for the seeds; because of this,
these large mammals are important in the early stages of forest
invasion of large expanses of dry forest pasture that are to be restor-
ed to forest. Santa Rosa National Park.
UPPER RIGHT. Red color morph of an adult tettigoniid grass-
hopper (katydid). The most common morph of this common spe-
cies is brilliant leaf green. Santa Rosa National Park.
LOWER RIGHT. Adult female agouti (Dasyprocta punctata),
a 3 kilogram prominent forest rodent that is a major dispersal agent
of large seeds in the dry forest. There is an acorn ( Quercus oleoides)
in her mouth. Dispersal of seeds by agoutis is critical the movement
of large pristine forest trees into the early stages of forest reinvasion
in dry forest. Photo W. Hallwachs. Santa Rosa National Park.
LOWER LEFT. Full-size caterpillar of the saturniid moth
Schausiella santarosensis. This large moth is found only within the
area of Guanacaste National Park, yet feeds only on the leaves of
guapinol (Hymenaea cuurbari/) which is a large legume tree that
ranges from Mexico to South America. Santa Rosa National Park.
 9

SUMMARY

Dry forest is the most endangered of the once widespread

habitat types in Mesoamericaitoday only 0,08 percent of the
original 550,000 km 2 is in preserves. This document describes
and discusses an $ l l .8 miilion prnject in northwestern Costa Rica
that will allow the dry forest organisms in Santa Rosa National Park
and on the evergreen-forested slopes of two nearby volcanos to reoc-
cupy the adjacent low-quality agricultural and pasture land. Simul-
taneously this project in tropical restoration ecology will have a
management focus designed to integrate the park itself, Guanacaste
National Parle, into Costa Rican local and national society as a
major new cultural resource in an area that is agriculturally rich but
culturally deprived. The 700 km 2 park will be large enough to
maintain healthy populations of all animals, plants and habitats that
are known to have originally occupied the site, and to contain
enough habitat replication to allow intensive use of some areas by
visitors and researchers. The biological technology for restoring a
large area of species-rich and habitat-rich tropical dry forest is
primarily fire control by managers, grass control by cattle, and tree
seed dispersal by wild and domestic animals (and as budgets permit,
intensive reforestation programs with native trees); this restoration
biology is already relatively well understood or currently being sub-
jected to field experiments. The sociological technology for integra-
tion of the park into Costa Rican society is straightforward educa-
tion of students and teachers at all ages and levels in the society,
and research on the biology of the park to obtain more information
to feed that education process. In addition to being a major cultural
resource, the park will have a variety of economic values such as
gene and seed banks for dry forest plants and animals, watershed
protection, reforestation examples and technology, ecotourism, and
conventional t<.>urism. The land to be incorponsted in Guamsca:,te
National Park is almost entirely owned as investment property by
people willing to sell it for a fair market price; $8.8 million is need-
ed for this purpose (S200 per ha, $81 per acre). A park that will
survive into perpetuity and display its cultural potential must have a
substantial endowment for technical and cultural management; a
minimum endowment of $3 million is needed for this purpose (an
operating budget of $300 000 per year). The entire project must be
in place by 1990, and about $1 million is needed immediately to
secure the habitats that are in danger of inmediate destruction.
•'
 11

INTRODUCTION

When the Spaniards arrived, there were 550000 km 2 of dry

forest on the Pacific side of lowland tropical Mesoamerica ( Figure 1).
Equal to about five Guatemalas i,n area, this dry forest occupied as
much or more of the Mesoamerican lowlands as did rainforest.
Today, less than 2%of this dry forest exists as relatively undisturbed
wildlands,and only 0.08%of it lies within nationalparks or other
kinds of conserved areas (Appendix 1). In contrast to today's con-
servation battle for the tropical rainforest, the dry forest conserva-
tion battle would have had to have been fought a hundred or more
years ago. To save what dry forest we still have, we are going to
have to give some land back to it. Habitat restoration is essential
before natural and anthropogenic fluctuations and perturbations
extinguish many of the small populations and habitat remnants that
have survived to this date.
Likewise, when the Spaniards a..--rivedthe dry forest habitat
was occupied by peoples with an intimate, if pragmatic, factual
knowledge and cultural understanding of the biology of dry forest.
Today, virtually all of the present-day occupants of the western
Mesoamerican pastures, fields and degraded forests are deaf, blind
and mute to the fragments of the rich biological and cultural heri-
tage that still occupies the shelves of the unused and unappreciated
library in which they reside. The schoolchildren of a Mesoamerican
town have neither their predecessors' contact with the natural
world nor the human cultural offerings of the large cities that are
supported by their parents' agricultural activities. What gives the
greater return - build a cultural center in the fields cut from the for-
est or lead the audience to the cultural center that already exists
in the forest? We must lead the audience to the forest, or all the
well-meaning conservation efforts in the tropical world will disap-
pear down humanity's throat.
Simultaneously, those tropical peoples that destroy their last
fragments of tropical forest close the door to one of humanity's
most ancient antl complex opportunities for cultural enlightenment.
Guanacaste National Park (Center Cover and Figure 2-4) has
three functions:

1° Use existing dry forest fragments as seed to restore about 700

km 2 of topographically diverse land to a dry forest that is suf-
ficiently large and diverse to maintain into perpetuity all ani-
12

mal and plant species, and their habitats, known to originally

occupy the site. It also must be large enough to contain some
habitat replicates that can absorb intense visitation and re-
search use. It will be the only such dry forest wildland area on
the Pacific side of Mesoamerica and is the only large area with
sufficient biological and social traits to become this. We have
the seed and the biological expertise; we lack control of the
terrain.
2° Restore and maintain a tropical wild/and so as to offer a menu
of material goods such as plant and animal gene banks and
stocking material, reforestation examples with native trees,
watershed protection, manipulation of vegetation by livestock,
recreation sites, tourism profits, wildlife management exam-
ples, agroforestry research data, educational programs (from
elementary levels to international symposia), and basic wild-
land biology data (which will in turn be part of the cultural
offering of Guanacaste National Park). We have the knowledge
and the interest, but we lack the arena and funds to develop it.
3° Use a tropical wild land as the stimulus and factual base for a
reawakening to the intellectual and cultural offerings of the
natural world; the audience will be local, national and inter-
national and the philosophy will be "user-friendly". We have
the audience, and we know what to start telling them, but
again, we lack the arena and the funds to develop it.

This document outlines and discusses the plan to achieve the

above three goals with Guanacaste National Park (GNP) as a new
national park in northwestern Costa Rica . The area contains
230 km 2 of established national parks and 470 km 2 of private land
(Figure 3-4, Appendix 4), and is about 1% of the area of Costa Rica.
One of the national parks (Santa Rosa) contains quite enough habi-
tats and populations to serve as the seed; they will be supplemented
by the population remnants throughout GNP and the pristine forest
remaining on the nearby volcanic slopes.
GNP is new in area and in concept from traditional Netropical
national parks. It is also old in area and concept in that it a) will
contain the well-established Santa Rosa National Park ( Center
Cover) and Murcielago National Park ( Figure 3-4), b) is foreshadow -
ed by 5-plus-year-old pilot studies in restoration ecology in Santa
Rosa, and c) is embedded in Costa Rican culture, a society that has
long held education and cultural development to be noble and legit-
imate human activities. In this society, disagreement traditionally
leads to debate rather than to physical attack .
The start-up and endowment cost is going to be $11.8 million
(US). This is the price tag on about 600 new Toyota jeeps in Costa
Rica or the cost of one medium-sized new biology building on a US
university campus . It is $4. 72 per person for each Costa Rica citizen.
The GNP plan is extremely site- and culture -specific. It is de-
signed to function in the exact context of the sparsely occupied
and low-quality pastures and degraded forests of a small part of
north -central Guanacaste Province, which is otherwise a rich agricul-
tural province. It must be evaluated in this context, and not in the
context of its appropriateness to other parts of the tropical world;
on the other hand, major fragments of the philosophy and technol -
 13

ogy underlying the plan are relevant to agroecosystem design

throughout the tropics, and much of the plan's design was stimulat-
ed by first-hand observation of the interaction of wildlands with
their societies throughout the tropics. The trials and examples of
GNP will be both training ground and models for the consolidation
and development of the other portions of the Costa Rican national
park system, and vice versa. The early stages of planning have been
developed in consultation with many persons and agencies inside
and outside of Costa Rica. Subsequent detailed planning of guide-
lines and their implementation will be conducted by committees of
interested persons and organizations primarily or entirely of Costa
Rican origin.
Within the next 5-10 years the wildland component of Costa
Rican society will be forever fixed in place; even worse, it is clear
that the Pacific coastal dry forest was destroyed faster and more
thoroughly than was the Atlantic rainforests (e.g., Figure 5). What
Costa Rican habitat is not in preserves will be dead, and the next
~tage (which we have already entered) is that of improving the qual-
ity of both wildland preserves and agriculture in the agroecosystem.
The preserves that do not become adequately integrated into Costa
Rican society will die as well. Small parts of the GNP plan are of
crisis urgency, and if the entire plan is not in place and functioning
by 1990 there will be no choice but to retreat to the 108 km 2 of
Santa Rosa National Park and apply the GNP plan on a scale that is
biologically and socially much inferior. If the terrain for GNP can-
not be purchased or otherwise frozen in its currently mildly damag-
ed and relatively unoccupied state within the next 1-3 years, like-
wise the plan will have to be abandoned for GNP and applied to
Santa Rosa alone.
This urgency comes about because the social and economic
stasis that has characterised the GNP area for the past 400 years
is at this moment coming to an abrupt end as a consequence of the
serendipitous coincidence of a) the obliteration of almost all pio-
neer agriculture in Costa Rica, b) the recent influx of outside influ-
ence from central Mesoamerica, c) the liquidation of family-land
holdings as owners pass retirement age, d) the corporatization of
the high quality farmland in the remainder of the province, and
e) the realization by large land owners that only a tiny fraction of
the GNP terrain is of agricultural use and that this use can only be
realized through labor-intensive farming by what amounts to human
draught animals. There is substantial risk that the current owners
will subdivide their large properties and sell the valuable parts as
luxury investment property and the other parts to gullible or des-
perate subsistence farmers. At the _present time, the entire 470 km 2
of GNP that is privately owned is supporting approximately 1200
head of cattle (though it could support perhaps five times as many
with intensive management) and a few ha of corn and sorghum.
Removing it from "production" will have no significant negative
impact on either the local or national economy.
The GNP plan outlined below follows the somewhat tradi-
tional format for conservation and land development plans. How-
ever, throughout there is the underlying philosophy that tropical
conservation has unwittingly used an incomplete recipe in its
adoption of national park and other conservation systems from
extra-tropical regions. It is traditional in, for example, the US -and
14

now in Costa Rica, to identify biologically important habitats,

obtain title to them, fence and patrol them, and view the task as
largely complete. Such an act is functional if society at large is pre-
programmed to recognize the jewel thus bestowed upon it as of
worth. If not, and this is the general case in .tropical conservation,
the story is only half-way through the first chapter of a long book.
The traditions of tropical conservation in general, and certainly
Costa Rica specifically, have to evolve with urgent haste to a mode
where the integration of the park into the social consciousness is
dominant and central to the entire plan. Those areas we view today
as endangered are probably already extinct and those areas we view
today as securely preserved are at best on the endangered list; they
will remain there until they are viewed in the same breath as schools,
churches, libraries and democratic government.
 ··......._.,___ ,-·-··-
\, .. U.S.A

'\
\ ''··,..) .""··-··,

~)

I
.......
.
~ ~\.
MEXlCO '\

i
CARIBBEAN SEA
~
r··-·· --f'''

···~ ·, ") I

i~""'"~.....
_.J
.•
•.

PACiFIC OCEAN

GUANACASTE NATIONAL

~TROPICAL DRY FOREST

Figure 1. The distribution of Mesoamerican Pacific dry forest (stippled) at

the time of the arrival of the Spanish conquistadores. Guanacaste National
Park is indicated in black in northwestern Costa Rica; the other preserved .....
V,
areas (see Appendix) are too small to be readily visible at this scale.
16

\
NICARAGUA

(
....... \ CARIBBEAN SEA
I r'
.)
/
.1·
~
. ....
, .......
) . .......
-· . .J -·-·-·, )
i

PACIFIC OCEAN

- Location of the proposed Guonocoste

Notional Pork in Costa Rico

COSTA RICA
Esc:1:2000000

Figure 2. The location of Guanacaste National Park in

northwestern Guana-
caste Province, Costa Rica. The Interamerican Highway passes through the
center of the Park, and the Park forms a continous swath from the volcano
topsto thecoast(seeCenterCover).
 PACIFIC OCEAN
/

---Proposed boundary
-;,,
•
0
_,...--..
..........
0
Santo Roso Notional Pork <Murcielogo annex>

Santo Roso Notional Pork

Guanacaste Volcanic Mountain Range

Forest Reserve-Velcano Orosi. PROPOSED AREA OF STUDY FOR
GUANACASTE NATIONAL PARK
Scale 1;200000
Rincon de lo Viejo Notional Pork

Proposed Areas to be included

Rural Settlerent of the Institute Of

Agrarian Development <IDA>

BAi'IA

Figure 3. The new area to be added to Santa Rosa National Park and Murcie-
lago National Park to form Guanacaste National Park (see also Appendix 4).
 as•30•

Cerro Casa Maritza

HOChO Hacienda ornsi

N
•.
'
,'' ·' '

.,.,.·...'l

'
,·:·

,;
I
, I

t
I , ••
,,--'.''
I
I
I
,ooos Rios

!, ....~.......
,-: Central Administration
Santa Rosa National Pork

''
}
\' i
_,
''
'' ·/ ''
,/
,', ',,, ....\
Boundary-Proposed Guonacaste National Park
,~
,~ ', ......
,
,..
,.}

C Cliffs / Exposed rock ,, '....

,,
~
'-,
,, '
~ Mangroves I ,'',,!'
w1----·
10°4!>'

[:=J Pastures and introduced grosses

~ Native grosses PROPOSED LANOUSE FOR
• D
~ Tropical dry forest in various stages of regeneration --\ , ..._,,/ GUANACASTE NAL.PARI<
.. Evergreen forest
~ Agriculture
Roads
lnteramerica Highway
o Sites of interest
FOTOINTERPRETACION: M.SC. Carlos Elizondo I.G. N.1986
D Airports D18UJO DEL MAPA: Bernot Murillo F.P.N.

Figure 4. Approximate location of the individual properties that are col-

lectively referred to in the text as Guanacaste National Park.
 1940 1950

Oceano Pacffico Oceano Pacffico

1961

Oceano Pacifico

1983

COSTA RICA
Cobcr1u..-a boscosa dcnsa
(mas de 80% de cobertura del
suclo) 1940-1983

l'lJArca boscosa dcnsa

Oceano Pacifico Fuente: OPSA, DGF.

Fi111ruI.

Colx:rlura boscosadcnsa (80-100°/o de cobc11uradcl suclo) •n Cosla Rica

en lusaiios 1940, 1950, 1961, 1977 y 1983
from Rodas 1985

Figure 5. The shrinkage in distribution of closed canopy forest in Costa Rica

from 1940 to the present. Only the volcano slopes within the area of Guana-
caste National Park (northwestern Costa Rica) contained enough pristine forest
to be recorded in this map in 1983 (Rodas I 985) . Note that the relative rate .
and thoroughness of forest removal 'has been substantially greater in the dry
western habitats than in the mountainous rainforest habitacs.
 19

THEREGION

IN GENERAL

Guanacaste National Park sweeps from the 1500 m peaks of Volcan

Oros{ and Volcan Cacao down to the Pacific Ocean, including the Santa Elena
Peninsula (Center Cover, Figure 2 -3) . It is thus a wide band across the north
central portion of Guanacaste Province and straddles the Interamerican High-
way. The southern boundary of GNP lies 30 km north of the Guanacaste pro-
vincial capital, Liberia (the population of Guanacaste Province is 200.2000)~
the northern boundary lies 30 km south of the Nicaraguan border at Penas
Blancas. The small fishing and farming village of Cuajiniquil is only a few
kilometers north of the northern boundary of GNP, and a land coloniza-
tion site of the Institute of Agrarian Development also lies on the north
central GNP boundary. The northern Guanacaste regional center of La Cruz
lies 15 km north of the northern boundary of GNP on the Interamerican
Highway. The town of Quebrada Grande and the growing villages of Potreri-
llos and Los Angeles lie a short distance from the southern boundaries. All of
these communities are based on agricultural land of much greater value than
that in GNP.
When the first conquistadores traveled from the present-day area of
Managua south-east to the Indian city of Nicoya (in the upper central Nicova
Peninsula) in the mid-l 520's, they passed within a few kilometers of the east-
ern edge of Santa Rosa National Park, at the center of GNP. In the late 1500's,
Hacienda Santa Rosa was established as a 700 km 2 beef-, hide- and mule-pro-
ducing ranch; GNP resides ahnost entirely within the boundaries of that origi-
nal hacienda, which was one of the very first to be established on land that is
today Costa Rica . During the following 400 years, Hacienda Santa Rosa was
variously subdivided into large pieces and sold to various owners, with numer-
ous changes of hand. The current pieces will be discussed in detail in a later
section.

ECOLOGICALPLACEMENT

GNP lies in the nearly continuous belt of what was once dry tropical
lowland forest from north of Mazatlan. Mexico to approximately the
Panama Canal in Panama (Figure 1). Pacific Mesoamerican dry forest ( e.g.,
Figure 6-8) is characterised by receiving 900 to 2400 mm of annual rainfall
during 5-7 months of the year (April-May to October-December) and no rain
during the 5-7 month dry season ( e.g. Appendix 3). The upper end of this
rainfall regime generates rainforest in certain other parts of the t ropics (e.g.,
Nigeria), but these other areas are not subjected to the strong winds that blow
during the first half of the dry season at GNP and are characteristic of much
 20

of western Mesoamericana. Southern Mesoamerican dry forests also have a

0-6 week short day season in the middle (July-August) of the rainy season; in
GNP, the timing and intensity of this dry season is extremely variable
( Appendix 3 ). While average values can be derived from weather data for the
GNP area, it is critical to recognize that the dry side of Mesoamerica is charac-
terised by 2-10 year series of exceptionally wet or dry years. These have the
effect of temporarily obliterating or reducing patches of fauna and flora in
the fine scale moisture mosaic. Under natural conditions , the sites of these
patches are reinvaded when the weather pattern changes . However, in small
dry forest preserves surrounded by agricultural land, there is no place from
which this reinvasion can occur.
Nocturnal low temperatures range from 16-23 C, and diurnal maxima
range from 26 to 38 C in most Mesoamerican lowland dry forest habitats;
GNP is not an exception. The dry season is substantially hotter than is the
rainy season, but the reverse seems true to humans because the dry season
winds create evaporative cooling.
In general, the lands once occupied by western Mesoamerican dry forest
have been converted to the pastures, breadbaskets and cotton fields of their
countries (e.g., Figure 9-11 J. Dry forest is easy to clear and maintain clear
with felling and fire; introduced African grasses (Figure JI) and cebu cattle
(Figure 12) give high yields from pastures; the dry season has somewhat of a
northern effect on soil nutrients and pests; the dry season allows easy access
and field preparation by machinery; grain crops grow as well in the rainy
season as they do in many extra-tropical summers; soils of ten receive the
downwind ash flow of the Central American volcanic chain; and the weather
is generally more cheerful than it is in the rainforest habitats on the Atlantic
side of Central America . Overall, the dry forest environment is relatively
similar to the tropical and extra-tropical habitats from which large-scale farm-
ing and ranching enterprises have been imported to Mesoamerica over the past
400 years. If rainforest were as easy to farm with extra-tropical agriculture
as is tropical dry forest, we would have no rainforest over which to argue
today.
It is commonplace to think of the Pacific Mesoamerican dry forests as
ecologically distinct and separate from the rainforests and upper elevation
forests of central and Atlantic Mesoamerica. However, recent studies of flying
animals in Santa Rosa and other parts of Guanacaste Province's dry forests
make it abundantly clear that many "rainforest" insects and some birds spend
the rainy season in the dry forest and the dry season in the rainforest or in
nearby moist forest refugia (e.g., Figure 24 ). Obliteration of either wet or dry
forest will obliterate these animals. One cannot view Costa Rica's national
park system as a series of islands but rather must view it as a network partly
connected by migrants. Some migrants can and do move hundreds of kilome-
ters (e.g., sphingid moths, birds) while for others, the moist refuge during the
dry season must be as close as a few hundred meters . GNP contains both
moist refugia and flyways between Guanacaste dry forest and rainforest on
the Atlantic side of Costa Rica.
It will be many years before we know what fraction of the "dry forest"
fauna has to have immediately adjacent evergreen forest (such as the cove
forest on Cerro El Hacha (Figure 24) and the evergreen forest on the two
volcanos (Figure 26-27) if it is to persist in the dry forest. What is, however,
abundantly clear is that these refuges are necessary if the dry forest fauna is
not to be severely reduced in species-richness (such as is presently encoun-
tered, for example, in the dry forests of the Mexican Yucatan Peninsula where
there are no moist dry season ref ugia owing to the highly permeable limestone
substrates) ...
 21

HABITATS
GNP consists of the Santa Elena Peninsula (85 million years old and
above the sea during that time; this is the oldest continually exposed surface
in Mesoamerica; Figure 17 lower), the Santa Rosa plateau (3-6 million year
old volcanic ash flow deposit; Cover), the ancient volcanic core known as
Cerro El Ha cha (Figure 24-25 ), the twin young volcanos of Oros{ and Cacao
(the most recent material being perhaps as young as 10,000 years; Figure
26-27), small areas of coastal marine deposits, and various alluvial fans eroded
off all the above substrates. Volcan Orosi and Volcfo Cacao are the most
northern and most isolated of the string of volcanos that extends south to
Volcan Turrialba, which is east of San Jose.
The original GNP vegetation contains a few to tens of km 2 of virtually
all kinds of dry forest habitat to be found over the broad latitudinal range of
Mesoamerican dry forest. It contains a complex mosaic. of the following
Holdridge Life Zones: Tropical Dry Forest, Tropical Dry Forest Moist Forest
Transition, Tropical Moist Forest, Premontane Wet Forest Basal Belt Transi-
tion, Premontane Wet Forest, Premontane Rainforest_, and Montane Rainfor-
est. The Islas Murcielagos and the tip of the Santa Elena Peninsula are proba-
bly the driest sites in the country. At its margins and interior, GNP has a varie-
ty of interfaces with coastal vegetation, river-margin vegetation and evergreen
rainforest. It has no natural lakes (but does contain seasonal swamps) and both
seasonally dry and everflowing rivers ( Figure 8, 13-15).
Owing to the diverse topography and geology of the GNP area, its many
habitats existed originally as a very complex mosaic. Today, these habitats
have been variously overlain and partly obliterated (and homogenized) by a
complex pattern_l)f cutting, burning, grazing and farming, followed by second-
ary succession ranging from O to 400 years in age. However, it is also clear
that somewhere within GNP lie minute to large patches of all the original
habitats and population fragments of all the plants and animals that were
present when the Spaniards arrived. The most pristine habitats lie in Santa
Rosa (Cover, foreground), in the ravines on the lower slopes of Cerro El
Hacha, on the upper slopes ( above 600 m) on the volcanos (Figure 26-2 7),
and in a few isolated patches up to a few tens of hectares scattered over the
remainder of GNP. The most seriously altered areas are the upper parts of
Cerro El Hacha ( Figure 25 upper), the Santa Elena Peninsula (including parts
of Murcielago National Park) ( Figure 17-18, 23 ), anct the wooded and brushy
pastures in all of the ranches to the east of the Interamerican Highway ( e.g.,
Figure 9,20).
Just as is the case with animals (to be discussed below), most GNP plant
species are widely distributed in the Neotropics. However, the widely distrib-
uted species tend to have distinctive dry forest populations (whether the
unique traits are genetic or ecological is unknown). Even the uniquely dry for-
est species are distributed widely throughout the Mesoamerican dry forest.
However, as is the case with animals. nearly all of these widespread species are
having their populations reduced to the tiny local populations in small reserves;
GNP will shortly be the home of an ever-growing list of Costa Rican "anthro-
pogenic endemics". GNP is also the only Costa Rican home of Ateleia herbert-
-smithii ( Figure 31 upper), the world's only wind-pollinted legume and the
tree that has become one of those selected to be widely distributed as a trop-
ical fuelwood species.
GNP's namesake is the guanacaste tree (Enterolobium cyclocarpum). It
is the national tree of Costa Rica and one of the best -known trees in Guana-
casteProvince(whichwasnamedafterthe tree).Ironically,this tree probably
did not occur naturally in Costa Rica in the period from 10,000 years ago to
22

when the Spaniards arrived, but was probably a more northern Mesoamerican
tree that came to Costa Rica as seeds riding in the guts of the first Spanish
horses and cattle. For the next 400 years it was distributed throughout
Guanacaste through seed dispersal by the horse and cow (just as it probably
was dispersed by the prehistoric giant mammals, including horses, that ranged
through Mesoamerica until 10,000 years ago. Today it is being extinguished
in many habitats through restriction and reduction of horse populations,
destruction of habitats by fire, and death of adult trees (through senescence
and lumbering).
The most prominent 15 dry forest habitatsin GNP are briefly character-
ised below:

1. SEASONAL (INTERMITTENT) RIVERS AND CREEKS ( Figure 8,

13}. During the dry season, all watercourses within GNP dry up except for a
few springs and the everflowing ones from evergreen forest (2 below). During
the rainy season, the amount and duration of flow in the seasonal water-
courses depends on the rainfall pattern. GNP's seasonal watercourses are
important dry season water sources (pools and springs), and the more ever-
green vegetation along the banks produces a cool and humid refuge as well.
The watercourses and watercourse banks are a major natural habitat for a
large fauna of ruderal plants and animals. The composition of this fauna and
flora, and the degree to which animals are dependent on a particular seasonal
watercourse depends strongly on its exact location, size, and rate of drying.
GNP will add one major seasonal river (Rio Potrero Grande, Figure 13) to the
three already protected in Santa Rosa, and numerous small ones (of which
there are very few in Santa Rosa).
2. EVERFLOWING RIVERS AND CREEKS ( Figure 14-15). The ever-
flowing rivers have their origin in the rainforested sides of the volcanos, and
then move out into the seasonally rain-free lowlands, generating linear dry
season oases. Such rivers are a major part of western Mesoamerican dry forest
ecology, but throughout the remainder of Costa Rica's dry forest habitats and
throughout most of western Mesoamerica they have been biologically obliter-
ated by deforestation, irrigation schemes and agrochemicals. In GNP, these
rivers (Rfo Centeno, Tempisquito, Gongora, San Josecito, Sapoa) contain a
unique t1ora and fauna (including fish and aquatic invertebrates that cannot
persist in the seasonally dry watercourses, but reinvade them each rainy
season from the everflowing rivers) , serve as major dry season refuges for ani-
mals, and have wet forest plants (on their banks) that do not otherwise occur
in the area. The existing dry forest parks, Santa Rosa and Murci~lago, do not
contain any everflowing rivers because they are topographically isolated from
the volcanos and too low in elevation. It is equally distressing that no extant
dry forest park in all of Mesoamerica contains an everflowing river system.
3. MANG ROVE SWAMPS ( Cover, Figure 16 ). The small estuarine
embayments along the coast from the southern boundary of Santa Rosa to
Cuajiniquil contain fine examples of dry Pacific coast mangroves. This habitat
has been generally destroyed over the past 200 years by bark (for commercial
tannin), post and firewood collectors along the Mesoamerican coast. However ,
the area of the mouth of the R1n Potrero Grandein Santa Elenacontainsthe
only pristine mangrove stand that occurs in northern Pacific Costa Rica.
4. DRY FOREST MARINE INTERTIDAL (Cover). Owing to inaccessi-
bility, the marine intertidal habitat is still relatively intact along the GNP
coast, in strong contrast to the remainder of northern Pacific Costa Rica
(where snail and clam collecting for food has all but eliminated most molluscs,
 23

for example). The turtle nesting beach is protected within Santa Rosa (Corne-
lius 1986) but if farmers were to colonize Santa Elena, the nesting beach
would be virtually impossible to protect from human egg gatherers and turtle
meat hunters. The five coastal preserves (Corcovado, Manuel Antonio, Cabo
Blanco, Ostional and GNP) would serve as an adequate Pacific coast national
seashore for Costa Rica.
5. ISLANDS. The Islas Murcielagos off the tip of the Santa Elena Penin-
sula (Figure 5) contain a perturbed but naturally severely depauperate dry
forest fauna and flora. In view of the decreasing rainfall gradient westward
along the Santa Elena Peninsula, and in view of the total absence of dry
season water on the islands, they are probably the driest terrestrial habitat in
the entire country. They have not yet been studied ecologically, but experi-
ence with other Pacific coastal Costa Rican islands suggests that they will be
found to contain very peculiar combinations of plants and animals, and may
have endemic populations (though not species). These islands are regularly
visited by fishermen and are being progressively deforested by anthropogemc
fires. Some, but not all of the islands still have enough of their original vegeta-
tion to be able to return to their original forest if protected from fire and
fire-wood collectors.
6. FRESH AND BRACKISH WATER SEASONAL MARSHES. These
marshes occur on the Santa Rosa plateau in the interior of Hacienda El Hacha
and Orosi, near the highway intersection at the northeastern comer of
Hacienda Santa Elena , and inland from the coast in the southern lowlands of
Santa Rosa. Small in area and severely disturbed by deforestation, fire and
cattle, these sites nevertheless contain a unique flora and fauna (e.g., Isoetes ,
L.D. Gomez, personal communication) which would likely recover its original
structure were it allowed to do so.
7. POST-MANGROVE PROSOPIS SWAMP. Immediately behind the
mangroves in Santa Rosa and a few places in Santa Elena and Murcielago are
unique patches of cacti , mesquite , divi-divi and other dry-land perennials. This
forest type has been obliterated by harvest of firewood (to be used in salt
extraction) in almost all other dry coastal Pacific sites in Costa Rica.
8. ALLUVIAL SEMI-DECIDUOUS BOTTOMLAND FOREST (Cover) .
Behind the coastal beaches were expanses of tens to hundreds of hectares of
flatland forest on rich and moist alluvial soil. They contained several hun-
dred species of trees, about 20% of which were evergreen. In Santa Rosa, as
well as elsewhere (e.g., Potrero Grande River valley bottom in Santa Elena),
these forests were severely but patchily felled ; however , within Santa Rosa a
mere 14 years of protection has allowed them to replace all fields and pasture
with 3-20 m tall secondary woody succession that contains the original animals
and plant species (though in very different proportions than originally). Small-
er versions of this forest occurred in Murcielago and behind other seasonal
river mouths in Hacienda Santa Elena.
9. STRONGLY DECIDUOUS HILLSIDE FOREST(Cover, Figure 17-18).
The sides of the Santa Rosa plateau, the hillsides of the Santa Elena Peninsula,
and the small slopes throughout GNP below 300 m elevation bear a complex
deciduous forest ranging from 2 m tall and totally deciduous in the dry season
(on south-facing upper slopes on ridges, especially on the peridotite or serpen-
tine substrates of the Santa Elena Peninsula) to 30 m tall with as many as half
of the trees evergreen. At least 600 species of broad-leaved plants occupy this
vegetation. A salient feature of this forest is that after it is cut, the woody
regeneration that appears in its place is much more deciduous than was the
original (until after the several hundred years that are necessary for the slow-
-growing evergreens to strongly reoccupy the site). Owing to the complicated
 24

disturbance regime over the past 400 years in the GNP area, extensive and
detailed study is required to know to what degree a particular patch of decid-
uous forest is pristine or a product of secondary succesion. Cutting and fire
has long ago cleared most of the deciduous hillside forest from most of the
Santa Elena peninsula (Figure 19), but small patches remain sprinkled over
the surface, patches that will spread and coalesce if the fires are stopped . A
peculiar and depauperate version of this forest occurs on a single minute
limestone hill in the Santa Rosa bottomlands.
10. EVERGREEN CANYON FOREST. The many escarpments and
small canyons of the Santa Rosa plateau bear (bore) a nearly evergreen forest
that was 30-plus m in height and dominated by guapinol (Hymenaea), tempis-
que (Mastichodendron), ojoche (Brosimum), terciopelo ( Sloanea), nispero
(Manilkara), caoba (Swietenia), guavo (Inga), higo (Ficus) and other large
evergreen trees lacking common names. These species also occur on the upper
slopes of the two volcanos, but intermixed with at least 100 other species of
trees that do not occur at the elevation of the Santa Rosa plateau. Just as with
the deciduous forest mentioned above, when this evergreen forest is cleared it
first regenerates as strongly deciduous secondary successional forest. The
shady and leafy evergreen canyon forests are extremely important local moist
refugia for animals of the deciduous forest during the dry season. GNP will
more than double the amount of this forest type under protection.
11. EVERGREEN OAK FOREST (Figure 20-22). The Santa Rosa plateau
(220-350 m elevation) and its extension to the base of the modem volcanos at
about 500 m elevation, was once covered with a nearly monospecific stand of
encino (Quercus oleoides) growing on a volcanic ash flow (rockhard substrate
with poor water retention and supporting only slow-growing plants). This
unique forest (it is the southernmost lowland oak in the Neotropics) extended
as far south as Bagaces and is the southern-most extension of what is known
in the US as Virginia live oak ( Quercus virginiana). Scattered throughout the
GNP oak forest are members of at least 80% of the deciduous and evergreen
forest species of plants; when the oak forest is cleared, they then take over the
site and convert it to deciduous or semi-evergreen forest. If the cleared site is
also burned, it changesto natural or introduced grasslandoccupied by the
most fire-resistant of the deciduous forest trees. If pristine or partly cleared
oak forest is protected from grass pasture fires, it very slowly reinvades the
site. However, while virtually all of Santa Rosa's oak forest is too seriously
perturbed to perpetuate itself, GNP contains at least five 5-20 ha patches of
essentially pristine oak forest, and several thousand hectares of only mildly
disturbed oak forest.
12. PASTURE HABITATS (Figure 12, 23, 25, 27, 31, 35, 36) Between
250 and 800 m elevation in GNP there are at least 200 km 2 of pasture (local-
ly termed sabanas or llanos). They are arranged in a complex network and
mosaic, and with many different histories. All GNP grasslands are maintained
as grasslands by anthropogenic fires every 1-3 years (Figure 30-31), most are
occupied by introduced African grasses, all had their origin in forest clearing,
and all begin to revert to woody vegetation as soon as the fires are stopped
(Figure 30 lower). The rate of reversion depends on grass species, soil type,
wind exposure, proximity of seed trees, pasture size, and wild and domestic
animal density as seed dispersers and grass suppressors (Figure 36). While at
least a quarter of GNP is now pasture , the configuration of the pastures and
their proximity to forest fragments is such that they revert rapidly to woody
vegetation; the process of this reversion is of great academic and applied inter-
est, and undergoing intensive field experimentation and analysis at Santa Rosa
at present .
 25

13. ATLANTIC-PACIFIC SEMI-EVERGREEN FOREST (Figure 26-28).

The broad erosion valleys and some slopes of the lower half of Cerro El Hacha
are still partly clothed in virgin forest. The forest is a checkerboard with 1-5
year old com and bean fields. Even so, it contains more standing tall virgin
dry forest than is contained in all Mesoamerican conserved areas put together.
One block of about 200 ha is the largest block of Mesoamerican tall virgin
dry forest in existence. The Cerro El Hacha forest is so evergreen that it
creates ever-flowing creek;s despite its six month rain-free season. It contains
enormous individuals of trees that are known as fence-post trees throughout
the remainder of Guanacaste (e.g., Gliricidia sepium) . The presence of
"Atlantic rainforest" plants and animals (e.g., the terciopelo, Bothrops
asper; dumb cane, Dieffenbachia) on Cerro El Hacha reinforces the impression
from Santa Rosa National Park that Guanacaste Province is now substantially
drier than it was when covered with its original forests. When cleared, the
Cerro El Hacha forest becomes grassland { e.g., Figure 24-25, 37) and its
streams stop flowing in the dry season. When the cut forest is allowed to
return to forest after a farming cycle, the vegetation is largely deciduous.
During the dry season, the Cerro El Hacha forest is extraordinarily rich in
insects that are obviously local migrants from the nearby dry forest.
14. VOLCANO SLOPE EVERGREEN RAINFOREST ( Figure 26-28).
From about 500 to 1000 m elevation on the western slopes of Volcan Orosi
and Volcan Cacao lies a nearly pristine rainforest that contains an amazing
number of Guanacaste dry forest species (but with much taller and more ever-
green life forms) as well as many species of the wetter portions of Costa Rica.
Likewise the animals in this forest are a mix of Atlantic and Pacific species; at
present we do not know which of the species from the dry forest are migrants
and which are residents). The extremely tall and large trees are very peculiar
in that they bear almost no vascular epiphytes and vines. This suggests that
the soil is moist but the air is dry. This habitat has at least a 7 month rainy
season, is 4-8 C cooler than is GNP as a whole (and therefore relatively more
moist), and displays much slower rates of forest invasioo into pasture than is
the case in GNP at lower elevations. This forest, and the semi-evergreen virgin
forest mentioned above, are major dry season refugia and corridors to Atlantic
rainforests for the many animals the pass the dry season away from the dry
forest.
15. CLOUD FOREST. The upper 500 m of elevation of 1500 m Volc_an
Oros{ and Volcan Cacao are bathed in clouds {Figure 26-27) at least 11
months of the year . The forest is dwarfed, heavily laden with lichens and
other non-vascular epiphytes, and drips continually. Its water is the starting
point for the everflowing rivers passing through the lower reaches of GNP.
Because the volcanos are very conical and pointed, these are the smallest
habitat islands of clou_d forest in Costa Rica, and those at the lowest elevation
( cloud forest normally starts above 1800 m elevation in Costa Rica). This
vegetationand its animalshaveneverbeen inventoried.
16. ATLANTIC RAINFOREST. Above about 600 m elevation, the east-
ern slopes of the two volcanos {Figure 3, Orosi Forest Reserve) are covered
with nearly intact rainforest. This forest blends gradually into the evergreen
forest on the western volcano sides { 14 above). Inclusion of this relatively
small area of rainforest in GNP is highly appropriate because it will maximize
the survival of the numerous populations whose members occur on both sides
of the volcanos. These are in turn essential to the survival of the populations
that occur only on the drier western sides of the volcanos and use the western
sides as moist refugia during the dry season.
26
GNPSIZE
Guanacaste National Park needs its large size for five biological reasons:
maintain habitat diversity, maintain adequate species population sizes, provide
dry season refugia and migration routes, minimize edge effects, and maintain
some replicated habitats for human park users.
I. MAINTAIN HABITAT DIVERSITY. Even a pristine "dry forest"
habitat is fraGtured into a mosaic of literally hundreds of kinds of tiny habi-
tats. This is because the physical and biotic diversity in slope, soil type, season-
al change in water flow, exposure to wind, bulk of vegetative cover, degree of
evergreeness, fire regime, rainfall pattern, etc. becomes magnified through its
impact on the amount and timing of water availability as the dry season
comes and goes. The scarcity of water during a tropical dry season is less
homogeneous than is the cold in a northern winter; the abundance of water in
a tropical rainforest obliterates many of the potential inter-habitat differences
that are so conspicuous in a tropical dry forest.
The high species richness of tropical dry forest is largely due to pooling
across the many different habitat types created by the heterogeneity described
above. This pooling occurs not only in the biologist's mind. Many species use
different habitats at different times of the year. A riparian tree may be polli-
nated by bats that at other times of year are visiting flowers on trees in open
upland dry sites. Many animals spend all or part of the dry season in a frag-
ment of evergreen forest understory and then move into the more resource-
-rich canopy of deciduous forest when the rains come; others, such as seed
weevils, may reproduce once per year in the dry season seeds of early succes -
sional herbs and then spend the rainy season hiding in rolled leaves in the
deciduous forest understory, waiting for next year's seed crop.
To accumulate a reasonable area of any one of the dry forest habitat
fragments, habitat fragments must be summed over hundreds of km 2 • Three
processes hamper the viewer's ability to see this:

(1) Until very recently, most research in the Costa Rican tropics was
done by visitors from extra-tropical regions; being largely from universities,
they visited during the northern summer, which is Costa Rica's rainy season.
In the rainy season, the dry forest is painted green and wet, and habitat
differences blur.
(2) Humans are accustomed to thinking in terms of vertebrates and .
large plants, and these are the most generalist organisms, the organisms least
likely to depend on very fine scale inter-habitat differences. The white-tailed
deer, collared peccary, jaguar, mountain lion, tapir, and white-faced monkey
may be encountered in all GNP habitats, albeit at different densities. How-
ever, the vast majority of the species in GNP are small -for example, there are
3,000-plus species of moths and butterflies and many more other species of
insects. Such animals show high habitat fidelity in where they breed, mate,
rest, etc. For example, if you want Bardaxima perses (a notodontid moth) in
your dry forest, you have to have a evergreen understory and it has to have
Ouratea lucens (Ochnaceae) shrubs for the caterpillars to eat. And so on and
so forth.
(3) Animals wander and plants are widely dispersed. This means that
habitats characteristically contain a large number of species that may best
be described as strays. This blurs habitat distinctiveness. On the other hand,
strays are also important parts of the food chain and pollinator and seed
disperser networks.
There is another reasvn why a dr y forest reserve must be large enough
to co11tai11i1iaJ1Y small replicates of habitats. Frnm year to year, dry forest is
 27

subjected to frequent and violent changes in weather. At Santa Rosa, for

example, the annual rainfall during the past five years has varied from 900 to
2400 mm of rain. The small dry season in the middle of the rainy season has
varied from O to 8 weeks in length. Habitats altered by these weather changes
recover largely through immigration from habitats and species pools in other
sites that were affected. In GNP, where the absolute number of habitats has
been severely reduced through habitat destruction, the problem will be even
greater until nearly total reforestation has been achieved. The scarcity of habi-
tat types is rarely reflected in collective terms like "dry forest". For example,
there are only 5 known ever-flowing springs in Santa Rosa's 108 km 2 • Santa
Rosa contains only two small canyons that are moist enough to maintain
vanilla orchids. In all of GNP there are only about 20 km 2 of habitat suitable
for the endemic legume tree A teleia herbert-smithii ( Figure 31 ). There is only
one pool in Santa Rosa large enough to serve as a dry season refugium of
muscovy ducks (but the everflowing rivers in GNP (Figure 14) will also serve
in that capacity if hunting is stopped). There is no patch of pristine oak forest
in Santa Rosa more than a few tens of meters on a side.
2. ADEQUATE SPECIES POPULATION SIZES. For large vertebrates
such as the jaguar, mountain lion, and tapir ( Figure 32), the breeding popula-
tion in Santa Rosa (10-50 individuals) is simply not large enough to avoid
inbreeding and subsequent genetic decay, genetic drift, and obliteration by
disease epidemics. The same applies to at least 30 species of dry forest trees in
Santa Rosa. Santa Rosa is not large enough to maintain even a single herd of
white-lipped peccaries . While insects and other small organisms would appear
to exist at population densities high enough that even a few km 2 of habitat
would be adequate to maintain them, in fact the past 5 years of intensive cen-
sus of moths at Santa Rosa has demonstrated enormous species-specific fluc-
tuations in density among years, with the species appearing to disappear at the
bottom of the fluctuation (e.g., Janzen 1984b). Likewise, small animals (and
plants) are often much more habitat -specific than are the large ones, with the
consequence that a much smaller proportion of the overall habitat is suitable
for them.
There is an important, but often overlooked, aspect of the loss of trop-
ical animals from a habitat. Almost all play conspicuous roles in internal
habitat structure through seed dispersal, seed predation, selective browsing,
pollination, predation on herbivores, etc. The biotic impact of the loss of
species is most dramatically displayed on islands, where whole suites of species
display demographies and behaviors grossly different from that of conspecifics
on nearby mainlands.
3. PROVIDE DRY SEASON REFUGIA AND MIGRATION ROUTES. A
substantial fraction of the dry forest animals use local moist areas as dry sea-
son refugia. Many of the mobile ones move as far as the semi-evergreen virgin
forest on Cerro El Hacha and the evergreen slopes of the volcanos ( up to 30 km
from the farthest point in GNP). It is likely that the final blow to the white-
-lipped peccary in Santa Rosa was the opening of the pastures along the Inter-
american Highway; these pastures form a broad unforested barrier between
Santa Rosa and the volcanos. Movements between the dry lowlands and moist
rainforest are not restricted to movements to escape the dry season, however.
It is clear that Santa Rosa is visited by some species of rainforest birds only
during the early dry season.
Strongly cross-tropical migratory species are also involved. For example,
at least 40 species of sphingid moths arrive in Santa Rosa at the beginning of
· the rainy season (from the rainforest), have one or two generations in Santa
Rosa, and then fly back over to the Atlantic side of Costa Rica to spend the
 28

remainder of the year ( e.g., Janzen 1984c) ..A dry forest preserve the size of
GNP is needed to maximize the survival of migration routes, maximize the
area of the breeding grounds for the rainforest species, and minimize the pos-
sibility that they will disappear because they cannot find a little dry forest dot
called Santa Rosa.
4. MINIMIZE EDGE EFFECTS. As a general rule of thumb, when wild-
lands connect abruptly with agriculturized land, edge effects in biological and
physical processes penetrate at least 1-2 km into the wildlands. Different ani-
mals and plants will experience this differentially, but at an absolute mini-
mum the habitats on 50-100 km 2 of GNP will suffer edge effects. These habi-
tats will be quite rich in vertebrates (owing to high productivity of vertebrate
food by secondary succession and edges). However, the blessing of increased wild-
life density is mixed. These animals then use nearby pristine vegetation more
heavily (browsing, fruit eating, trampling), and disperse many more secondary
successional seeds in and into it than is normal. Even with all the protection
that Santa Rosa receives, for example, this process in strongly altering the
small pieces of pristine forest within the park(Janzen 1983a) The concept of
small blocks of pristine forest in the neotropics is simply an optical and tem-
porary illusion .
5. HABITAT REPLICATION FOR HUMAN USE. A user-friendly na-
tional park must havea varietyof areasand habitatsthat are freelyopen to
moderate to heavy public educational and recreational use. An area sufficient
for this purpose is likely to be considerably larger than the area required sim-
ply for traditional biological reasons. Humans have an impact, whether they
are individual researchers, school groups, tourists or solitary hikers; complex
tropical ecosystems are easily perturbed by human presence and there must
be enough habitat replicates that some can be used by humans without fear of
eliminating a unique habitat. Likewise, some major research projects may re-
quire the relatively exclusive use of a particular habitat piece for many years.
Finally, long-term manipulative reforestation model projects will require
substantial space. GNP is large enough to contain small to moderate numbers
of replicates of at least some of its more spectacular but fragile habitats (e.g.,
everflowing rivers, beaches, evergreen canyon forests, mangrove forests, pris-
tine forest of all kinds, xeric ridges, springs). It also contains sufficient area
for replicated substantial natural and manipulative reforestation projects.

FAUNA
Of the area to be included in GNP, only Santa Rosa National Park has
detailed faunistic surveys to date. Its 7 50 species of plants sustain at least
17 5 species of birds, 115 species of mammals, 3 140 species of moths and
butterflies, and at least 10,000 other species of organisms. Extrapolating from
preliminary visual surveys of the remainder of GNP and from surveys of other
parts of Costa Rica, the birds of GNP should be about 300 species. the mam-
mals about 140 specie.s, the moths and butterflies about 5000 species, and the
plants about 3000 species when all of GNP is surveyed. Most of this increase is
due to the inclusion of the semi-evergreen virgin forest on Cerro El Hacha and
the western sides of the volcanos. If these estimates err, they err on the low
side.
The GNP fauna is overall representative of that of dry forest throughout
Pacific Mesoamerica. It contains many wide-ranging species that also range
into rainforest arul into South America. There is, however, an abundant dis-
tinctive dry forest fauna that is found, in Costa Rica and elsewhere, only in
the dry forest. When a GNP faunal list of a major group such as birds, moths,
 29

bats, or beetles is compared with one from a Costa Rican Atlantic rainforest
there is only a 10-20% reduction in species richness. Jhis reduction is · so
small because there are many dry forest species that do not occur in the rain-
forest; the latter category substantially lengthens the GNP species list. Owing
to the extreme seasonality of GNP, one might expect that its species richness
would not be great in comparison with extra-tropical seasonal habitats. For
many groups, however, this is not the case. There are more species of butter-
flies, large moths, and mammals in GNP's 700 km 2 than in all of the US east
of the Mississippi River.
Many animal life forms classically thought of as "rainforest animals"
(e.g., sloths, tapirs, white-lipped peccaries, spider monkeys, howler monkeys,
white-faced monkeys, army ants, morpho buttherflies, scarlet macaws, toucans,
red-lored parrots, carnivorous bats, etc.) occur in GNP but at lower density or
only as seasonal members of certain habitats. GNP does not receive a heavy
dose of extra-tropical migrant birds (though dry forest does do so in other
parts of Mesoamerica); these birds appear to make more use of Costa Rica's
evergreen rainforests than her dry forests. Furthermore, the northern migrants
leave Costa Rica for extra-tropical regions about the time (or earlier) that the
rainy season begins and the large flush of food appears during the first two
months of the rainy season.
Along with the many wide-ranging species that occupy Santa Rosa there
are a very few endemic species (e.g., the saturniid moth Schausiella santaro-
sensis and see Cover). However, many of the dry forest species that once occu-
pied all of the Costa Rican dry lowlands are having their populations dramati-
cally reduced to tiny populations in widely scattered preserves such as GNP,
thereby rendering them "anthropogenic endemics". In addition, many of the
less mobile animal species in GNP's dry forest belong to a population that is
morphologically distinct from the same species on the wet side of Costa Rica.
In general, GNP individual birds, moths, and monkeys are smaller and lighter
in color than are their rainforest conspecifics. We do not yet know how much
of this difference is genetic and how much an ecological expression of the
shorter rainy season, longer dry season, greater insolation, greater temperatu-
res, and other seasonal forces.
The GNP fauna is conspicuous in that it reinvades abandoned pasture
vegetation more rapidly than occurs in analogous habitats in Costa Rican rain-
forests. The same is true for the woody vegetation, and the two are mutualis-
tically related. The animals move seeds as well as eat the fruits and foliage.
There is also a distinct gradient within GNP; pasture invasion by forest is
much more rapid in the central and western parts of GNP {drier, warmer and
lower elevation) than it is on the slopes of the volcanos (moister and cooler).
 31

HUMAN OCCUPATION
OF GUANACASTE NATIONAL PARK

PREHISTORIC

The GNP area overall has been at best trivially surveyed or developed for
its archaeological sites. Santa Rosa contains a variety of unstudied ancient
grave-sites as well as at least on~ very large village site in the lowlands near the
ocean. The headwaters of the Rfo Sapoa on the lower slopes of Cerro El Hacha
have been thoroughly studied and related to Indian groups living slightly more
to the north. The recent spectacular results from intensive archaelogical explo-
ration of the Tilaran region (at the elevation of the volcanic slopes in GNP) 80
km to the southeast suggest that there may be still much of value to be under-
stood about the site's archaelogy.

CONTEMPORARYOWNERSHIP

Land ownership of GNP is almost entirely in the form of large holdings

(Figure 4) managed as business investment and owned by persons living else-
where. At least in 1986, seven owners of large properties, one owner of a
small property, and one collective colony of settlers on small parcels are the
people with which direct negotiations are necessary. In addition, several tiny
land fragments need to be obtained from large ranches on the south boundary
of Santa Rosa. The ownership of each of the parts of GNP is described below
in detail. The relationships with settlers and ranchers living near the bounda-
ries of GNP will be discussed later.
I . SANT A ROSA NATIONAL PARK (SANTA ROSA SECTION).
(Cover). 108 km 2 • On 27 June 1966, SRNP was expropriated and declared a
National Monument (Law No. 3694). By Executive Decree No. 1562-A of
20 March 1971 it was declared a National Park. On 4 May 1977, Santa Rosa
was enlargedby ExecutiveDecreeNo. 7013-Aso that the park's major drain-
age basins were almost completely enclosed by the park . Santa Rosa is occu-
pied by a small staff of about 20 administrators and rangers, about ten of
which are in the park at one time; all of them have homes elsewhere in Costa
Rica.
2. SANTA ROSA NATIONAL PARK (MURCIELAGO SECTION), 122
km 2 • On 13 November 1980, Hacienda Murcielago was expropriated and
established as an addition to Santa Rosa National Park by Executive Decree
No. 12062-A.LawNo. 6794 of 25 August1982ratifiedboth sections-Santa
Rosa and Murcielago - as Santa Rosa National Park.
Murcielago is occupied by a tiny staff of about 4 administrators and
rangers, all of whom have homes elsewhere in Costa Rica.
3. ISLAS MURCIBLAGOS. About 3 km2 • These multiple smallislands
off of the tip of the Santa Elena Peninsula (Figure 3) belong to the Costa Rican
32

government and are in the process of being officially declared part of Santa
Rosa National Park. They are unoccupied but are frequently used as rest stops
by fishermen from Cuajiniquil.
4. HACIENDA SANTA ELENA (Figu,re 12, 17, 23). About 130 km 2 •
Santa Elena occupies the area between Santa Rosa and Murcielago on the
north and south, and the Pacific and the Interamerican Highway on the west
and east. Santa Elena is apparently owned as investment property by the Odol
Corporation in the United States. It is currently undergoing infrastructure
development (roads, airport, buildings), annually subject to free-running wild-
fires that then threaten Santa Rosa and enter Murcielago, and lightly grazed
by cattle. It is occupied by a Costa Rican overseer with a few helpers and their
families (headquarters near the Interamerican Highway). There are irregulari-
ties in locations of the fences between Santa Elena and Santa Rosa National
Park, but these will be unimportant if GNP can incorporate Santa Elena.
5. CERRO EL HACHA (Figure 24-25). About 50 km 2 • The north and
northeast portion of Cerro El Hacha is part of Hacienda El Amo/El Hacha/ A-
guas Buenas/Guitarra belonging to Sr. Luis Roberto Gallegos and other large
ranches, while the southern and southestern portion belongs to the Colonia, a
collection of small farms occupied by about 16 owners since I 980 and coming
originally from the area of Santa Elena and Monteverde (Puntarenas Province).
All owners are willing to discuss sale of their respective portions of Cerro El
Hacha. While Sr. Gallegos recognizes the watershed value of Cerro El Hacha
for the remainder of his cattle ranch holdings, the farm owners are in the pro-
cess of clearing the forest to grow 1-2 corn or bean crops and "improve" the
land value. The Colonia has already cleared approximately one third of the
unique forest on Cerro El Hacha and will destroy much of the remainder in
the 1987 and 1988 dry seasons
6. HACIENDA EL HACHA DE RANCHOS HORIZONTES ( Figure 20
background). About 40 km 2 . This investment property is owned by Mr.
Cecil Hylton of the US and managed by Sr. Gustavo Echeverri of Ranchos
Horizontes, an agricultural corporation operating out of Liberia. At present,
El Hacha is operated as a minimum density cattle ranch. It is occupied by
about 2 administrators and their families.
7. HACIENDA OROSI ( Figure 26). About 30 km 2 • This investment pro-
perty has the same ownership as does Hacienda El Hacha de Ranchos Horizon-
tes. At present Orosi has had almost all of its cattle removed and is occupied
by 1 administrator and his family (at the ancient Orosi ranchhouse). Mr. Hyl-
ton has very kindly agreed to donate Hacienda Oros1, piece by piece, to the
Nature Conservancy as part of GNP. Sr. Echeverri has promised no further
development and that GNP may begin patrolling Hacienda Oros{ to prohibit
hunting and other intrusions ( this patrolling begins in March, 1986 ).
8. OROSI FOREST RESERVE. 105 km 2 • The portions ofVolcan Oros{
and Volc;n Cacao above about 550 m elevation (Figure 26-27) are govern-
ment forest reserves and cannot be legally cleared of forest. There is even a
questionable law (Ley 1917, 195 5) that declares the area within 2 km of the
volcano craters as a national park (Bonilla 1983). The land ownership, how-
ever, is still in the hands of private individuals (e.g., portions of Hacienda Oro-
si, Hacienda Centeno and Hacienda San Josecito are within the Orosi Forest
Reserve). At the present time, almost no one lives within the Orosi Forest
Reserve on the west, north and east sides of the volcanos, but settlement has
crept well past the margin of the Orosi Forest Reserve on the sou them flank
of Volcan Cacao. While the Reserve is legally protected, in fact it is gradually
beingclearedbecauseregulationsarenot enforced.
9. HACIENDA POCO SOL (Figure. 8, 9, 20, 22, 26). About 40 km 2 .
 33

This operating cattle ranch has been in the Burgos family for at least 40 years,
but the owner, Sr. Mario Burgos, lives in San Jos.eand is willing to sell the pro-
perty for fair market value. Sr. Burgos has kindly promised, in deference
to GNP, to do no development modification of Poca Sol during 1986 (but he
will continue with · his development planning). His son, Sr. Gustavo Burgos,
lives on the property and manages it, along with his other agricultural proper-
ties in Guanacaste. There are about three administrative families and several
ranch helpers living at the Ranch Headquarters near the Interamerican High-
way. In local terminology, Hacienda Poca Sol consists of two properties
known as Poco Sol and Garzal. A newly constructed Voice of America trans-
mission station occupies a few hectares of Poco Sol near the Highway (Figure
20).
10. HACIENDA CENTENO (Figure 15). About 40 km2 • This investment
property is owned by Mr. Gene Peacock, a US citizen resident in San Jose. It
consists of three properties, Centeno, Guancastillo and Mata Redonda; the
latter is the most interior and on the slopes of Volca.n Cacao. Mr. Peacock
plans to lease Centeno as cattle grazing land to neighboring ranchers, and has
plans to develop the river bank alluvium for coconut orchards and the ever-
flowing river for snail ponds. However, he has kindly agreed to stop develo.l'"'.
ment for 1986 in deference to GNP. He will consider sale of the entire
Hacienda for a fair market value. Hacienda Centeno is occupied by one admi-
nistrator and his family.
11. HACIENDA SAN JOSECITO (Figure 27). About 30 kin 2 • This pro-
perty has been in the Baltodano family since 1935 and is currently owned by
Sr. Aristides Baltodano of San Jose. Sr. Baltodano is eager to sell San Josecito
and is currently receiving offers from other individuals; however. he is attrac-
ted to the idea of having it end up in GNP. He does not plan development dur-
ing 1986. San Josecito is currently occupied by one administrator and his
family.
12. HACIENDA TEMPISQUITO (Figure 14). About 15 km 2 is of
interest to GNP. This property has also been in the Baltodano family since
1935 and is currently owned by Sr. Jorge Baltodano of Liberia. Sr. Baltodano
is willing to consider selling the semi-forested portion of the northern part of
Hacienda Tempisquito, leaving the ranch headquarters near the Interamerican
Highway in his hands. He does not plan development of the area of most
interest to GNP in 1986. Hacienda Tempisquito has two administrators and
their families.
13. FIN CA JENNY (Figure9, 21 ). 4 km 2 . This small piece of investment
property is owned by the Gulf Land Company of Sra. Jenny Perez of San
Jose. It was carved out of the corner of Hacienda Santa Rosa more than 200
years ago as a real estate scheme. Sra. Perez is willing to sell Finca Jenny, but
is currently asking a price roughly double its market value. This small piece of
relatively intact forest is critical to the biological integrity of the largest and
deepest evergreen canyon forest (Quebrada Puercos) in Santa Rosa National
Park. Finca Jenny is occupied by an administrator and his family.
14. FINCA GUAPOTE. About 2 km 2 • The site is a tiny corner of Finca
Guapote which is in turn owned by a very large cattle ranch, Hacienda Ahoga-
dos, alon~ the southern boundary of Santa Rosa National Park. The site con-
tains a large spring that is an important dry season watering site for animals
from the park; Hacienda Ahogados prohibits hunting in Finca Guapote, but
the prohibition is only partly effective because it is at the extreme northern
boundary of the Hacienda . This site and Finca Jenny combined will seal off
the Quebrada Puercos canyon forest from outside threat and intrusion. The
possibility of sale of the site to GNP by Hacienda Ahogados is being investi-
gated at present. No one lives at _the site.
 34

15. HACIENDA ROSA MARIA (Figure JO). About 3 km 2 is of interest

to GNP. The site is a strip of sorghum and cotton fields along the southern
boundary of Santa Rosa National Park. While almost all of Hacienda Rosa
Marfa drains to the southeast (Rio Tempisque drainage to the Gulf of Nico-
ya). a small border area drains into Santa Rosa (Pacific drainage) and poses
an imminent and serious threat to the finest ·of the large seasonally dry rivers
in the park (Ri'o Poza Salada); agrochemical and silt drainage from these
fields has already destroyed ( 1984) a major creek system within Santa Rosa.
The owner is Sr. Pedro Abreu of Miami, and the Hacienda is managed by his
son, Sr. Carlos Abreu of San Jose'. They have agreed to help with avoiding
pesticide contamination for the time being, with the understanding that in
the final negotiations over sale of this tiny fraction of Rosa Marfa to GNP,
there is discussion of the possibility of connecting Hacienda Rosa Maria to the
Santa Rosa electricity line. No one lives on the site under consideration,
though a ranchhouse with one administrator and family is nearby.
16. SOUTHWEST MARGIN OF SANTA ROSA. About 10 km 2 • While
presently unthreatened, the southwestern corner of Santa Rosa was estab-
lished through rough terrain and unbroken dry forest without consideration
of the drainage details. This minute area has yet to be explored in conjunction
with the Santa Rosa neighbors. No one lives at the site.

HUMANRESOURCESIN THEAREA
Whileoverlappingin capabilities,inclinationsand potential, three some-
what distinct groups of human resources are already present in GNP and its
inmediate vicinity.
1. RESIDENTS. A large number of people living in the GNP region
(roughly La Cruz to Liberia, and the small town areas of Cuajiniquil and Que-
brada Grande) have residence roots 2 or more generations in length. Many of
these people have grown up with minimal formal schooling (though all are
literate) but have lived a varied life rich in the details of survival where farm -
ing, ranching, fishing timber extraction, civil service, and small business are
the primary occupations (hunting has largely been extinguished along with
the game). The overall social structure is Spanish/European/US/modem to the
extent that resources permit. Upward mobility is minimal and therefore indi-
viduals with strong mental and psychological ability are encountered at sub-
stantially lower income levels than would be the case were native ability to
strongly determine an individual's economic level and social status. Town and
country residents display very strong curiosity about anyone or anything that
approximates a learning experience, remember copious amounts of material
and instructions without writing them down, and leap on opportunities to
better their material goods.
The residents around GNP (e.g., Figure 29) form an obvious and unex-
ploited knowledge and labor pool for the day-to-day management of GNP.
They already know how to carry out most of the technical aspects - fighting
fires, placing fences, maintaining horses as riding and pack animals, maintain -
ing trails and buildings, herding cattle, identifying and understanding vegeta-
tion and trees, dealing with biotic challenges (snakes, ticks, diseases, thirst,
hunger, wounds, etc .), etc. They learn rapidly about vehicles if the are not
already familiar with them. If they know it is part of their job, they are self-
-motivated to do these things. However, they need training in the facts of
biology (a combination of organizing the biological miscellanea they have
already accumulated and teaching them major biological facts), in how to tell
biological (sensu latu) stories to others, and in having the self-confidence to
 35

somewhat aggressively guide others through a learning routine. The major focus
of park managers drawn from this pool will be on the interface between the
users of GNP and GNP biology, though these managers will also have basic
maintenance responsibilities. These will be minimized through the enactment
of the principle that the park interior will largely take care of itself; if labor-
-intensive manipulation is required for a research or reforestation program,
that labor will largely be provided by the program itself.
A minimum number of 50 well-trained and apprentice residents will be
needed to manage GNP in the early stages. These people will have to live in
or immediately adjacent to GNP , on homesteads that will belong to GNP (if
they are inside GNP) but allow individual initiative in gardens and milk cows,
and in house modification and upkeep. It is clear that some of them will be
drawn from the personnel already managing the various haciendas in GNP
(Figure 29 right) while others will come from nearby farms and the towns of
Cuajiniquil, La Cruz, Liberia, etc. ( Figure 29 left). The GNP resident managers
will be maintained permanently in GNP and have individualized responsibili-
ties. They will be sufficiently unisolated that their children have access to
schools and the family has access to a normal social life.
It is assumed that certain local residents will sufficiently excel in the
challenge outlined above that they will climb through the GNP administrative
structure. Likewise , it is likely that some will find research and teaching activi-
ties to be sufficiently interesting and rewarding to use them to move into
those worlds, either within or outside of the GNP area.
2. COST A RICAN VISITING MANAGERS. Costa Rican managing visitors
to GNP will range from students from other parts of the country who come to
participate in a research/teaching program or do their own research/teaching,
to technical advisors that are temporary parts of the GNP managing staff.
Some of these may stay on as part of the resident managing staff, but it is
assumed that they will then become residents of the area. Such persons will
oft en bring specific important skills with them, but will require training in the
technical and philosophical peculiarities of living and working in the GNP
area , and in the art of making the park maximally user-friendly.
3. FOREIGN VISITING MANAGERS. Foreign visiting managers will be
largely research scientists and research students. While they conduct their own
studies they will also be active participants in the development of the user-
_friendly status of GNP. Their contribution will include aggressively making
their studies well-known to the resident managers, collecting and providing
background data on what organisms are in GNP and on their natural history,
being advisors for Costa Rican apprentices in field biology, aiding in planning
specific management programs (including the development of the tourism
value of the park) , and giving public lectures on their research at GNP in other
Costa Rican institutions as well as in their home societies.
-
 37

Figure 6. ( Upper). Semi-deciduous dry forest in the middle of the dry

season (March); the very dark tree crowns in the center are evergreen guapinol
courbari/J. (Lower).The samesemi-deciduousdry forest as above,
( Hyme11aea
but in the middle of the rainy season(July). NatureTrail,Casonaarea, Santa
Rosa.
 38

Figure 7. (Left). Same forest as in Fig. 6 during the dry season. but from
the interior , looking up at the Monument be hind the Casana. ( Right) . Sam e
forest and view as on left. but during the rainy season.
 39

Figur e 8. (Left). Quebrada Pitahaya , a seasonal watercourse, in the early

dry season (January) . (Right) . Same view of Quebrada Pitahaya during a
rainy period in the late rainy season (November). Near the Interamerican
Highway.HaciendaPocoSol.
40

Figure 9. (Left) . The Interamerican Highway at the east end of Santa Rosa
National Park. The Park entrance is at the middle of the long diagonal section;
the two elongate pastures at right center (in the Park) are several hundred
years old and cut out of oak forest. The area below the Highway (lower
left) is patchily disturbed oak forest in Hacienda Poco Sol. Finca Jenny lies
at the upper center (to the right of the severe curve in the highway) and
contains much of the forest in its vicinity. The thoroughly deforested
pasturelands to the south of the Park are evident at the top of the photograph.
(Right). Rice fields and other representative farm and pasture land in the
Liberia area. This thoroughly deforested habitat has only remnant large trees
and almost no reproduction by large forest trees. It is also lacks almost all
forest vertebrates and insects .
 41

Figure 10. View northwest across Santa Rosa National Park from about 300
m elevation over Hacienda Rosa Mana (Santa Elena mountains on the back-
ground horizon) . The uniform gray fields in the foreground are unharvested
cotton. The pale jaragua pastures in the background (in the Park) are intermixed
with deciduous forest patches of various ages. The cotton fields adjoin directly
with the Park's unused jaragua pastures.
42

Figure 11. View to the north from the eastern central part of Santa Rosa
National Park. Volcan Orosf is under the clouds in the background, and an
ungrazed and unburnedjaragua pasture (Llano Guacimal) lies in the foreground.
Pastures such as these can be eliminated by stopping fires, moderate grazing
by cattle, and allowingwild vertebrate seed dispersers to persist at natural
density.
 43

Figure 12. (Upper). Representative cebu cattle in a herbaceous pasture from

which thev have eliminated almost all grass by their grazing. Hacienda Santa
Elena. (Lower). Free-ranging horses at Laguna Escondida in Santa Rosa
National Park. Both tire cattle and horses are important dispersersof forest
tree seeds into large expanses of grassland, and important in reducing the
grasses that compete with tree seedlings.
44

Figure 13. Lower Rio Potrero G rand e. a seasonally dry river passing through
semi-evergreen and deciduous forest during the dry season. When the upper
drainage basin of such a river (Figure 18, lower ) is deforested. it thoroughly
dries out during the dry season; if the original forest cover is retained (e.g ..
Figure 17, upper), the upper riverbed has pools that last through the dry
season.
 45

Figure 14. Rio Tempisquito (upper Rio Tempisque) , an everflowing river,

wher e it passes through Hacienda Tempisquito. During the dry season, almost
all of the water in this river originates in the evergreen forests on the sides of
Yolcan Oros1and Volcan Cacao. Such rivers are missingfrom Santa Rosa
National Park . but were characteristic of the Guanacaste lowlands to the
south of Guanacaste National Park.
46

Figure 15. E.io Centeno, an everflowing tributary of the Rfo Tempisquito

(Fi~ue 14) in HaciendaCenteno. In addition to being unique for their ever-
flowing water, such rivers in Guanacaste National Park are also unique in
being free of agrochernical contamination and being totally unstudied.
 47

Figure 16. The tea mangrove, Pelliciera rhizophorae, growing in the back
portion of the mangrove swamps at the mouth of the Rio Potrero Grande
in Hacienda Santa Elena. In northern Guanacaste Province, this species of
tree is known only from this site.
48

Figure 17. (Upper) . The upper drainage basin of the Rio Nisperal , as viewed
from the north toward Playa Naranjo in Santa Rosa National Park. This
deciduous forest once covered all of the dry hills of the Santa Elena Peninsu -
la, and will be a major source of animals and plants to restore Hacienda Santa
Elena as part of GuanacasteNational Park. (Lower). The upper drainage basin
of the Rfo Potrero Grande in Hacienda Santa Elena, or seasonally dry river
lying adjacent to the Rio Nisperal above. These deforested hills were once
covered with the same forest type as in the upper photograph. The upper and
lower photographs were taken from the same site.
 49

Figure 18. Dry season deciduous forest on the lower slopes behind the coastal
plain in Santa RosaNationalPark (the same vegetationtype as in the posterior
part of Figure 17, upper). The native columnar cactus ( Lemaireocerus aragonii)
lives in a seasonally available desert.
50

Figure 19. (Upper) . Deforested hills on the sides of the upper valley of the
Rfo Potrero Grande (Figure 17, lower). This deforested state is maintained
purely by frres. (Lower) . Deforested hills on the sides of the valley at th e
mouth of the Rio Potrero Grande; on the left, a grass fire has burned upwind
to the ravine in the photograph center but failed to cross the rocky and
relatively grass-free ravine bottom . Such heterogeneity of burning · pattern
creates heterogeneity in rates of forest regeneration and kind of forest type
to appearon a site.
 51

Figure 20. View to the north from over the inland center of Guanacaste
National Park. Hacienda Poco Sol is directly below and grades into Hacienda
El Hacha in the upper center. Hacienda Santa Elena is to the left of the
highway on the left. The electric power transmission line passes through on
the right. The Voice of America transmitting station is to the left of lower
center. Almost all forest in the photograph is mildly to badly perturbed oak
forest.
52

Figure 21 . A 300 -plus-year -old oak (Quercus oleoides) in a remnant of

pristine oak forest along the interior margin of Finca Jenny . Such trees
occur in pristine forest patches only in areas outside of Santa Rosa National
Park,thoughthere are a few individualsremainingin the Park.
 53

Figure 22. Soil and rock erosion of an old road on white volcanic ash soils
where oak forest once stood in interior HaciendaPoco Sol.
54

Figure 23. Upper hilltop grass pastur es in central Hacienda Santa E lena.
These native grass pastures are maintained by annual fires but were once
coveredwitha 2-6m talldeciduousforest.
 ))

Figure 24. Cerro El Hacha as viewed from the lower slopes of Volcan Oros1
(looking northwest). The nearly totally deforested (and annually burned)
upper slopes stand in sharp contrast to the.._somewhat sheltered ravines contain-
ing remnants of pristine semi-evergreen forests that are important dry season
moist refugia for dry forest insects.
56

Figure 25. (Upper). Top of Cerro El Hacha (Figure 24 ), currently covered

only with native grasses but once forested . (Lower). One of the few remaining
forested upper slopes of Cerro El Hacha in the process of being deforested .
On the right is a recently cleared corn field. in the center and top left is
intact forest , and on the left and foreground is pasture cleared of forest
sometime in the past several hundred years.
 57

Figure 26. (Upper). Volcan Oros{ as viewed from the center of Hacienda
Poco Sol (Volcan Cacao to the right). The foreground was once covered
with oak forest and still has a few rer.mant patches. (Lower). Volcan Orosi'
as viewed from Hacienda Orosf. The pastures cut out of the lower ·volcanic
slope pristine forest are only 20-30 years of age, but will return to forest
very slowly.
58

Figure 27. (Upper) . Volcan Cacao as viewed from above Hacienda Tempis-
quito (Volcan Oros1 to the left). The heavily disturbed forest in the fore -
ground was a mosaic of oak and deciduous forest , grading into the evergreen
forest on the lower volcanic slopes . (Lower). The sinuous elongate pasture
on the right slope in the photograph above. It is assumed that the southern
boundary of Guanacaste National Park will pass along the spine of the ridge
down this pasture, or slightly.to the right of it. As on Volcan Orosf,such
upper elevation pastures (400-800 m) return to forest only very slowly as
compared with those of lower elevations.
 59

Figure 28. A view through the pristine evergreen forest canopy on the western
slopes of Volcan Oros1. This tree is 40 m tall, and like the other trees in this
forest, very free of vines and vascular epiphytes.
60

Figure29. (Left). A research assistant, Sr. Roberto Espinosa, in Santa Rosa

National Park . He comes from Cuajiniquil , has spent his life living out the
challenges in this area, and is responsible for the execution of a variety of
complex biological research tasks in the Park . Such a person will play a major
role in both Park biological instruction and generation of information about
the Park. (Right). The caretaker of Hacienda Orosi, Sr. Mateo Mata. He comes
from the area, is very competentat the daily tasksof runninga dry forest
cattle ranch, and is the kind of person who would constitute a major part of
the management of Guanacaste National Park.
 61

Figure 30. (Upper). Aerial View of natural pasture reforestation in Santa

Rosa National Park after five years without fire (center left, and see below).
The paved road serves as a firebreak, protecting the experimental area from
the recently burned control area ( center right and see Figure 3 1). On the far
right is regularily burned jaragua pasture that had not yet been burned at the
time of this photograph. (Lower) . The experimental jaragua pasture mentioned
above that has been protected from fire for five years . Almost all the broad-
leafed_ p~a(l.tsare seedlings and saplings of large forest trees, and almost all
are wind-dispersed ..
62

Figure 31. (Upper). The jaragua pasture control for the experimental plot
in Figure 30 (lower). This grassstand is burned annually.The singlesurviving
tree is A te/eia herbert-smithii. (Lower). The above control plot after its
annualfire.
 63

'I I 111I
0
II Q 1 1 l

Figure 32. (Upper). The tapir (Tapirus bairdii), an important seed dispersal
agent in Guanacaste National Park . This relative of the horse does not live
in open pastures, but crosses them and therefore sometimes defecates in
them. (Lower) . All of these seeds were in a single defecation of a wild tapir
in Santa Rosa National Park. The seeds are of ce1:,~_ero (Pithecel/obium
saman), a major timber tree of Guanacaste dry forest. A cenizero fruit 19 cm
in length is included at the top for scale. The seeds on the -left are dormant
and viable, those on the right were killed by germination 'lnd digestion in the
tapir, and those in the center germinated shortly afte r being defecated.
 64

Figure 33. (Upper). The coati (Nasua narica), an important seed dispersal
agent in Guanacaste National Park. This relative of the racoon eats many
fruits and defecates the seeds in a viable state . (Lower) . A pile of coati dung
containing over a hundred viable seeds of Styrax argentea, a rare evergreen
dry forest tree in Santa Rosa National Park.
 65

Figure 34. A small cluster of the winged wind-dispersed seeds of mahogany

or caoba (Swietenia macrophyl/a). In Santa Rosa National Park, the dry
seasonwindscarry these seedsas far as 200 m intoabandoned
gra~spastures,
where their seedlings are among the numerous species of wind-dispersed
seedsto first invadesmallpastures.
 66

Figure 35. Posts isolated in the center of a several-hundred hectare jaragua

pasture. Birds flying across such pastures stop at such posts and defecate
seeds upon resuming flight, resulting in an accumulation of tree seeds in the
area of a post. These seeds produce small forest nuclei that gradually spread
and coalesce into continuous forest if not burned.
 67

Figure 36. (Upper). A horse-grazed jaragua grass pasture in Santa Rosa

National Park during the rainy season. Such an open habitat is ideal for
tree seedling establishment if seed dispersal occurs into the area and the
site is neither burned nor cleared with a machete . (Lower) . The same pasture
as in the upper photograph, but on the other side of the fence . The dense
jaragua stand is a severe competitor for tree seedlings and offers a heavy
fuel load for grass fires in the dry season; such fires are hot enough to destroy
the above ground parts of almost all woody plants .•
 C\
00

Figure 37. A small remnant patch of forest on the pastureland on the lower
slopes of Cerro El Hacha. The pasture is maintained by fire and occupied only
by woody prants that are extremely resistant to fire. In moist years, the forest
patch expands and on dry years it contracts due to penetration by the annual
fires.
 69

THE ACTION PLAN

As will become evident below, we already have the biological knowledge

that is needed to make GNP a reality. There is already an audience for its pro-
ducts and that audience will grow. There is no doubt that people can be found
with a deep interest in carrying out GNP's working operation. What we do not
have is the money to purchase the terrain or for the endowment that will
generate the management funds.
A . ALLOW FOREST REINV ASION

If all fire and livestock were deleted from GNP today, and the site simply
allowed to revert to its own vegetation, the grass patches of less than 5:10 ha
would be largely woody vegetation within 20 years while the largest expanses
of pasture {e. g., in the Santa Elena Peninsula, Figure 17, 19, 23, 24) will
require 50-200 years to attain this status. Dry forest populations and habi-
tats will immediately begin to return to their orig_inal sizes and areas. The
en tire area will require at least 100-1000 years to begin to approximate the
full structure of pristine dry forest. As will become evident below, some of
the forest reinvasion processes can be substantially speeded up by habitat
manipulation, and this will be done in GNP as resources permit .
Below we briefly summarize the biological and managerial aspects of the
forest reinvasion process at GNP. Again, as mentioned earlier, it is important
to note that these processes, and especially their rates, will be different in
other parts of the tropics (and outside of them).

I .-FIRE. WITHOUT A SUCCESSFULFIRE CONTROLPROGRAM,

GNP WILL CONTINUE DOWN THE TRAIL .
TO ALMOST PURE GRASSLAND

Fire will be the single largest threat to GNP for decades. Furthermore, in
those GNP habitats that are too fragile to allow cattle grazing as a way to
depress grass density , fire will be even a greater threat after GNP formation
than before; in a single growing season an ungrazed GNP pasture generates
enough grass fuel to carry a fire hot enough to kill all aboveground woody
small plants and sublethally damage the large trees. The most dangerous grass
stands are unbroken (ungrazed) 2 m tall dense swards of jaragua (Hyparrhenia
rufa), the introduced African grass {e.g., Figure 3JJ. However, even the lower
and less dense native grass pastures in Santa Elena {Figure 23) are a fire threat
when not grazed. On the other hand, when the rues are stopped the grass pas-
tures rapidly fill with seedlings and saplings of large trees.
Fires in Guanacaste's dry forest are not "forest fires" in the sense of the
70

popular imagination. They either burn in grassland and consume both the
grass and woody vegetation, or they bum through the litter layer underneath
an established forest. Such fires are easily extinguished by backfiring from
previously burned fuel-free lanes or beating out (especially at night). A con-
tinuous problem with fires in gi:_asslandoand dry forest mixes is that the fire
ignites old logs and standing dead trees, and these continue to burn and gener-
ate burning cinders that are blown across fire lanes for days afterwards. The
past five years of fire control (and lack of it for the past 14 years) at Santa
Rosa make it clear that the technology of fire elimination is feasible and
straightforward; the problems lie in the social problem of insuring that the
technology is applied year after year without fail.
All fires in the GNP area are anthropogenic. Lightning does not occur
during the dry season and when it occurs at the beginning of the rainy season
it is accompanied by rain. GNP fires (as are Santa Rosa and Murcielago fires)
are of two kinds: those that start in surrounding ranchland and burn or blow
into the park, and those started by humans within the park. There is absolute-
ly no circumstantial or biological evidence that natural fires were ever part of
the Guanacaste dry forest environment (e.g., Figure 25, 27). However, some
forest edge destruction may have occurred through fires escaping from Indians
burning secondary succession in preparation for planting.
All incoming fires can be stopped by the simple procedure of burning a
100-200 m wide fire lane along all park boundaries that have grass on either
(or both) sides. For GNP, this will be about 30 km of fenceline. Ideally, the
fire lanes are burned on pastureland belonging to neighbors. The fire lane is
set by mowing two 3-6 m wide parallel strips 100-200 m apart in the first
month of the dry season (late November to early December) and burning
these mown strips at a time of day when the standing grass is too moist to
carry a fire. About 2 months later the wide strip between the fire lanes is
burned (preferably at night). The annual fire lanes must be burned in the same
place each year, resulting in a strip free of dead tree trunks. In the case that a
fire burning toward the park is moving fast downwind in the daytime, a back-
fire may also be started from the wide fire lane.
In addition to the above lanes, strategic narrow fire lanes must be cut
and burned such that they partition the park into major blocks of grassland-
-forest mix with the long axis across the wind. These blocks serve in combat -
ing fires that begin within the park either by being blown in or from accidents.
Fires within the park are combated most effectively by getting to them inme -
diately while still small, and both backfiring and directly beating out the fire .
This requires rapid location of the fire and rapid mobilization of a maximum
number of persons to fight it. If treated properly, such fires rarely consume
more than a few ha of vegetation. GNP fires are smoke-rich and can be located
easily from a high point if a fire watch is maintained during the dry months.
GNP is oriented such that the long axis points upwind and the eastern-
most end is sealed with unburnable evergreen forest. This will render the fire-
breaks along the northern and southern boundaries especially effective. Addi-
tionally, intensification of agriculture in the areas to the south and north of
GNP will lower the incidence of dry season fires as ever more land is shifted
from pasture to cropland and more ranchers realize that fires damages most
types of pasture.
An occasional fire will probably escape or invade GNP. Does it matter?
Each time a grassy or young wooded area burns, it further postpones the day
when the vegetation will have returned far enough to forest to be essentially
unbumable.
 71

A few words on the biology of Guanacaste fires are in order here:

a. When a cinder blows into intact dry forest and ignites a snag or log, that
log normally burns up on the spot without creating a fire that spreads
through the litter; the living deciduous woody plants do not sustain a
fire in dry forest. However, if such a fire occurs in the early afternoon
late in the dry season, deciduous forest litter may continue to burn
slowly along at ground level until evening_humidity increases and falling
temperatures extinguish it.
b. When a grass fire burns downwind into a dry forest, the heat is sufficient
to kill and incinerate marginal trees (and saplings) and the wind carries
the fire tens of meters into the forest. An outcome is that when a pasture
is cut out of dry forest, the annual fires cause the pasture to move down-
wind.
c. The later it occurs in the dry season, the more thorough is a fire's incin-
eration of patches of woody succession and isolated trees, and the less
likely it is to bum around moist swales and creek banks (and see Figure
19) .
d. While creeping litter fires appear to do little more than kill the occasional
sapling, they do severe cryptic damage. As the fire burns the litter
accumulated against large tree bases, the heat kills the cambium in small
areas that are not visible at the time. If another ground fire passes within
the 5-20 years that are required for the tree to grow over this wound, the
fire finds ready access to the tree core thro!Jgh the dead area and the
large living tree is cut off at the base. The same process occurs in pastures,
where trees that are sufficiently heat-tolerant to survive for decades in
the light grass fires of grazed pastures have their bases damaged and then
cut off by the intensive fires of ungrazed accumulated fuel.
e. When a small fire does occur in dry forest (and especially, in young
secondary successional dry forest), it kills a small area of saplings and
overhead tree crowns. This allows more light at ground level in the fol-
lowing rainy season, and the site grows a dense ground cover of herbs
and grasses. When this material is added to the dead woody stems from
the previous fire, it makes the site particulary susceptible to a strong
local fire that creates an enlarging and grass-choked hole in the forest.
Subsequent burning creates a rapidly growing brushy pasture ( e.g., Uhl
and Buschbacher 1985). This process, associated with selective lumber-
ing and road and trail penetration, has been a major process for conver-
sion of dry forest to brushy pasture in Guanacaste during the past 400
years . When cattle are added to the habitat, the process is accentuated
or retarded , depending on details of season, stocking rates, pasture
quality, fire frequency, etc.
f. While fires kill aboveground parts of woody plants, many dry forest
species freely sucker sprout from roots and stumps. A 2 m tall sucker
shoot may belong to a root system that is hundreds of years old, and
such a sucker shoot often again grows into a large tree when the fires
are stopped.
g. In GNP habitats, so much grass accumulates in a single rainy season in
the absence of livestock that its fire kills all aboveground woody small
plants and many trees. So-called "controlled burning" to depress fuel
levels is disastrous.
h. When the fires are stopped , the return to woody vegetation can be rapid
even in the absence of livestock to depress competing grasses. In the
oldest experiment in Santa Rosa, five growing seasons have been suf-
72

ficient to convert a 200-year-old 4 ha jaragua grass pasture to a rapidly

closing stand of young trees (Figure 30-31 ).

2._:SEED MOVEMENT AND SEEDLING ESTABLISHMENT

Seed dispersal and seedling establishment is under intensive study at

Santa Rosa and all of the last 14 years of observations and experiments are
directly pertinent to forest restoration in GNP. There is a rich fauna of wild
animals in dry forest that move seeds into pastures, fields and woody succes-
sion (Figure 32-33). These animals, and their consequences, are much more
evident in Costa Rican dry than in wet forest habitats (though this difference
inay not be real because this kind of succession has yet to be studied in Costa
Rica in a rainforest area where the animals are sufficiently protected to occur
at natural densities). The ingredients of greatest importance are distance to
seed sources, seed dispersal mode, interdigitation of pasture and forest patches,
and species and number of animals. Once the seeds have arrived, their ability
to generate forest depends on the traits of the soil and pasture grass.
The first wave of woody succession into GNP pastures ( Figure 30) has a
very high proportion of wind-dispersed trees (e.g., Figure 3 4) on the down-
wind side of forest. Such seeds are moved in > abundance up to 200 m by the
dry season winds. However, such wind-dispersed plants do not readily get to
the centers of pastures of hundreds of ha in area (though they could very
easily be scattered there by hand).
Trees with animal-dispersed seeds, on the other hand, have much more
diverse patterns of input to pastures and other open areas. Large seeds (as well
as small) are swallowed in the forest and then defecated or regurgitated at
various distances out into pastures by the animals crossing or using them ( deer,
peccaries, coatis, tapirs, cows, horses, coyotes) (Figure 32-33). The behavior
of these animals tends to concentrate defecated seeds along ravines, at rock
outcrops, near isolated trees and in other sites potentially protected from fires
and desiccation. Seeds defecated in stream beds in forest by these animals are
also carried into open areas far from the forest.
Forest birds and bats do not readily venture into the expanses of large
pastures, but they do cross smaller pastures and often temporarily perch in
isolated trees (just as mammals walking through pastures often pass or pause
beneath such trees) (Figure 35 ). Before or while flying off, these birds defecate.
The consequence is that the growing island of woody vegetation that begins
to apoear around lame isolated oasture trees is usually entirely made up of
species that are animal-dispersed. This process emphasizes the importance of
the appearance of isolated large trees in large pastures; in GNP, such trees are
often guanacaste (Enterolobium cyclocarpum) trees and cenizero (Pithecello-
bium saman) trees., both of which are dispersed by horses and cattle.
Once seeds have arrived in GNP pasture (and other kinds of old field
habitats), their primary challenge is the 1-2 m tall dense stand of grass that
blocks sunlight, collects nutrients, and physically blocks growth. Livestock at
moderate to low density encourage woody succession into pastures by reduc-
ing the amount of grass (Figure 36). They do eat some woody plants but most
species are ignored unless grass is in short supply. The resultant woody succes-
sion has a somewhat different species structure than does succession without
livestock. The use of livestock in grass depression and succession management
in GNP will be highly controlled, and terminated as sites reach a stage where
the grass is no longer seriously threatening the woody succession through
competition and irre.
 73

3. INTENSIVELY MANAGED REFORESTATION

Left to itself without fire, GNP will revert to forest, and do it more rap-
idly if the pastures are manipu lated with livestock. However, GNP contains
suff ciently large areas of grass that it can fill an important educational role by
explicity generating forest types with certain compositions that are desired by
the agroforestry community. These experiments should be large enough to
serve as significant models and placed strategically to aid in dissecting the lar-
gest blocks of grassland in GNP. For example, it would be technically easy to
establish wide strip forests of fast- and slow-growing timber species on the
downwind margins of major traditional firebreaks, thereby eventually elimi-
nating the need for the maintenance of the firebreaks. Such a mixed forest
might well, for example, be composed of cedro ( Cedrela odorata), caoba
(Swietenia macrophylla), pochote (Bombacopsis quinatum), guanacaste
(Enterolobium cyclocarpum). cenizero (Pithecellobium saman), guapinol
( Hymenaea coubaril}, n1spero (Manilkara chicle) and tempisque (Masticho-
dendron capiri). These trees are all native to GNP, part of the natural second-
ary succession in GNP, and widely recognized in Costa Rica and elsewhere as
valuable timber trees. They range from fast-growing light-weight timber to
extremely slow-growing and dense timber. Much is already known of the biol-
ogy of these trees in the GNP area, quite enough to begin experiments as soon
as land is available.
The labor and other costs of such intensive land management within
GNP will not be provided by the regular managerial staff of GNP, but rather
will appear as explicit research programs within other budgets. The same
applies to harvest, care and manipulation of the natural seed and gene bank
that GNP obviously is.
GNP is not the place for the explicit introduction of "valuable" exotics.
We are already paying,a huge price for one such - jaragua. The last thing we
need is to have to try to eradicate eucalyptus, melaleuca, Australian acacias,
and other such useful trees. The same apolies to introduction of wild "useful"
animals. The indigenous dry forest flora is rich in species with the useful
properties of exotics plus many other useful traits. In like manner, it is
imperative that the indigenous dry forest plant and animal gene pools in
GNP be kept as pristine as possible. Trees and animals introduced even from
other parts of Costa Rica represent a serious genetic threat, to say nothing of
the diseases and parasites they carry. The release of confiscated pet wild
animals into the GNP area must also be halted.

4. CESSATION OF HUNTING

Poaching in Santa Rosa is presently a trivial problem, but there is a

serious problem with hunting in the remainder of GNP. The wild mammals
and large birds are important not only for their own sake, but because they
are major seed dispersal agents. Santa Rosa contains a naturally high (but
heterogeneous) density of peccaries, deer, agoutis, pacas, monkeys, tapirs,
coatis, coyotes, bats, guans. curassows. and other seed dispersers: these ani-
mals are at severely endangered densities in the remainder of GNP. This is due
to both hunting in the past and present and to habitat modification. However,
they are sufficiently mobile that with protection they will again attain their
natural densities in GNP.
The mammal most threatened with extinction in Costa Rica today is the
white-lipped peccary (Tayassu pecari). A herd is believed to still be resident in
the upper rainforest on the volcano slopes, and in January 1986, a small herd
74

of 31 animals was encountered while it was passing through Santa Rosa (W.
Hallwachs, personal communication). While a reforested GNP is large enough
to support one or even two white-lipped peccary herds, there is only a small
chance that hunters will allow a herd to survive in the general area long
enough for this to come about.
The major hunting in the GNP area is by pleasure hunters from La Cruz,
Liberi<!,, and San Jose, rather than by rural hunters desperate for meat. Ces-
sation of this hunting requires three things. First, the GNP managerial staff will
be strategically placed, and there will have to be selective vehicle checks at
key places. Second, and much more important, the community of pleasure
hunters will be subject to an intensive and personal education campaign by
GNP biologists. Third, the GNP staff will have to be trained out of the atti-
tude that they are highway patrolmen and that the loss of an occasional deer
is a serious threat to their egos.
In addition, there is some local hunting for meat in the area that will
become GNP. It is clear that much, if not all of this hunting can be stopped
by directed education at the elementary school level; children will be among
GNP's best ambassadors. Additionally, some of the better hunters are likely to
end up on the GNP managerial staff. As with the pleasure hunters, the loss of
an occasional deer or collared peccary to a local meat hunter is trivial compar-
ed to the potentialimpactof an arrestedand bitter poacherduringthe period
that it takes to educate the local population away from hunting within GNP.

B. ORGANIZATIONAL INFRASTRUCTURE

GNP will be organized and run under an elaborate constitution that is

explicitly designed for it and embedded in Costa Rican public law and decree.
Neither current traditions nor current national park laws are adequate to
guide the complex interactions necessary between a society and a national
park as a cultural institution.
The constitution will be the output of one or more national and interna-
tional workshops held in GNP, and attended by interested parties and
representatives of all relevant Costa Rican institutions and organizations.
These workshops will be organized and conducted by a consortium of the
foundations and government organizations in Costa Rica that are most directly
interested in the maintenance and survival of GNP (e.g., National Park Service,
National Park Foundation, Fundacion Neotropica, CA TIE, Wildlife Service,
Forest Service, Institute for Agricultural Development, Institute for Tourism,
Guanacaste Province Government, University of Costa Rica, Universidad
Nacional, Ministry of Planning, Ministry of Education, Ministry of Agriculture,
etc)
GNP will be the exclusive property of the National Park Service of Costa
Rica and a small portion of its budget will be derived from the NPS budget.
However, the bulk of its budget will be derived from the investment revenue
generated by the Guanacaste National Park Endowment Fund. This fund is
currently within the Costa Rica Program of the International Program of the
Nature Conservancy, but will be transferred to the National Park Foundation
of Costa Rica as GNP becomes a reality.
The directorship of GNP will be guided by the GNP constitution and
answer to the National Park Service and to a relatively small executive com-
mittee made up of those with most direct interest in GNP function and
survival. The director will reside in or adjacent to GNP and be a Costa Rican
citizen, as will the other full-time employees of GNP. While GNP administra-
tors at all levels may participate in training and organization at other Costa
 75

Rican national parks, they will not be transferred to them unless they wish to
be transferred.

C. USE PROGRAMS

All three primary goals of GNP center on the use and relevance of GNP
to the people that live outside the park, from local to global. Many details
of making GNP maximally user-friendly will be formalized in the park's
constitution; others will be invented as circumstances arise. However, there
are a few major subjects that can be briefly mentioned here since their
inclusionis a certainty.Whenwe state that GNP must be maximallyuser-
friendly, it must be recognized that there are many kinds of users . Furthermore,
if GNP fails to make the Costa Rican population fully aware and understand -
ing of its presence, even its short-term success as a traditional biological
preserve will be very short-lived.

1. INFORMATIONSTORAGE
There will be thorough documentation of GNP vegetation with aerial
photographs and ground truthing so as to have a reference point for the
multitude of regeneration experiments that will be automatically set in
motion simply by establishing the park . Where possible, animals will also be
censused. In addition, an elaborate set of records of management regimes,
experiments, errors, and other perturbations will be kept both on-site (in a
building with no fire and humidity risk) and in some distant protected place.
All information will be available to all interested parties (though commercial
users will be expected to make appropriate contributions to the GNP
Endowment Fund) . Rapid and detailed publication will be encouraged for
observations, experiments and results from GNP. Journalists, science writers,
educators and others wishing to write about GNP information will be
encouraged.
Where habitat manipulation has occurred, the experiments themselves
are a form of living information storage and their protection will be maintained
to perpetuity (as will their records).
2. INVENTORY
While certain groups of organisms are fairly well known for Santa Rosa,
we are woefully ignorant of just what organisms live in GNP and where.
Inventory surveys of flora and fauna are desperately needed and will be
encouraged as a contribution from the world of taxonomists. Likewise, the
taxonomic status of Central American organisms is sufficiently poorly
developed that it is imperative that GNP specimens are widely circulated in
the world's taxonomic centers such that revisionary work is certain to include
GNP materials. Finally, the arduous task of providing basic field guides and
reference collections to the tens of thousands of GNP species must begin,
group by group. As researchers, we cannot come to understand what holds
GNP together without names for the units in the matrix. What is equally
important, but too little appreciated, is that we cannot bring the biological
stories of GNP to the external audience without having names for the
organisms. These names allow not not only local reference, but also allow
us to connect what we find out about GNP to what is known elsewhere.
3. RESEARCH
Research within and about GNP is a critical aspect of its development
76

as a cultural institution. You have to do more than take people to the

symphony; you have to have something to play for them. Furthermore, if
GNP is to realize its many pragmatic biological functions (gene bank, seed
bank, reforestation, etc.), there must be active research programs within
GNP. There are many small and very cheap ways to make a tropical area
maximally attractive to field researchers, ranging from streamlined administra-
tion of red tape to meals at cost to erecting spacious primitive dwellings. The
GNP area is often a more foreign environment to Costa Rican researchers than
it is to foreign field biologists, and active steps will have to be taken to change
this as well. GNP can easily become a model meeting ground for researchers
from different cultures but with interests in problems in common.
It is traditional for tropical research results to move out. of the tropics
into the common knowledge (courses, journals, symposia) of extra-tropical
countries, and then trickle slowly down the educational ladder and back into
the tropics through courses taught much later to students in the tropics.
Direct participation by Costa Ricans in research projects within GNP, first
as technicians and apprentices and later as principal investigators, has great
potential to short-circuit this lengthy process .

4. ACCESS
All points within GNP are accessible, though sometimes only after
considerable effort. by some combination of vehicle, horse and/or foot
travel. However, to make GNP maximally user-friendly, a strong system of
trails, seasonal roads, and all-weather roads will have to be established. The
Interamerican Highway, cutting through the center of GNP. is an ideal starting
point for many kinds of access and public education. Properly signposted and
with forest regenerated to its sides, it will not be a serious barrier to animal
movements . Guanacaste Province. rich in road building activity and mechanized
agriculture, is not poor in the machinery needed for road development within
GNP; what is lacking at present are the connections to mobilize this machinery
once GNP is a rality.
5. USE ZONING

GNP will be heavily used by people, and these people will sometimes
have conflicting interests as well as pose potential threats to some aspects of
park biolo_gy. Thf evolving management constitution for GNP must contain
a detailed and broad-minded zoning system for various uses, and this must
be developed with not Only GNP's biological peculiarities in mind, but with
strong consideration given to GNP's development as an educational institution
and intellectual stimulus. As mentioned earlier with respect to the size of
GNP, quality use zoning will be greatly augmented by the presence of habitat
replicates.
6. EDUCATIONALFACILITIES
GNP must be developed as an outdoor , living educational institution.
In addition to the traditional services of extensive educational centers rich
in displays and printed information, and the traditional abundantly signposted
nature trails, there must be a strong ability and availability within the manage-
ment personnel to serve as educational guides . Costa Rican society is very
oriented toward verbal communication; this makes education more labor-
intensive, but also allows it to be more tailor-made for particular audiences.
The written material appropriate for a group from the University of Costa
 77

Rica is not likewise appropriate for an elementary school group from Cuaji-
niquil.
Perhaps the greatest amount of educational return for the smallest
intellectual and cash output within GNP (and even Santa Rosa at present),
would be the development of a cheap scheduled truck that serves as a reliable
bus, complete with a driver with a minimal understanding of habitat locations
and what they off er of biological interest.
However, among the most important educational facilities for GNP will
be several individuals with the primary responsibility of serving as field
biology teachers at large. They must circulate among the schools, high schools,
technical schools and the branch campus of the University of Costa Rica in
the GNP area, and provide illustrated lectures on the kinds of biology in GNP.
They must give other public lectures and serve as prominent guides when
there are "open house" days at the park (e.g., Guanacaste Day on 25 July;
Santa Rosa Anniversary Day on 20 March). They must be available as know-
ledgeable biological guides within the park, as well as be aggressively involved
in training park guards to be both good biologists and good teachers in the
field. They will be essentially ambassadors for GNP, and their knowledge of
both established GNP biology and current research programs will have to be
extensive . On a geographically more distant basis, it will be important that
GNP research and development programs be prominently represented in
international research and educational symposia (and especially those held
in the tropics).
Simultaneously GNP must aggressively introject its presence into the
contemporary efforts by the Costa Rican Open University and other organiza-
tions to increase the teaching ability of school teachers in biological subjects.
This must include not only traditional written materials and lectures in courses,
but organized field trips to GNP designed to aid school teachers to understand
the rich educational material offered by a national park. There is also a
growing awareness in Costa Rica of the value of collaborative seminar series
and courses among the four university-level institutions; GNP must be both
a contributing participant and occasionally the host in such activities. While
Santa Rosa is already visited occasionally by field trips from the universities,
CATIE and the Organization for Tropical Studies, there has been almost no
aggressive sale of the cultural offerings of a site like GNP.
The tourists, be they from other Rarts of Costa Rica or international,
will obviously benefit directly from the development of GNP as an educational
as well as a recreational, research, etc. institution. However, it is important
that GNP become more than a simple stop along a tourist route. This will
require some imaginative activity in developing tourist living facilities within
and near GNP. It is also assumed that private individuals in the GNP area, as
well as in more distant places, will develop their own guiding and other
tourist services as the opportunity presents itself. The staff and planning of
GNP must reach out aggressively to interact with the growing ecotourism
infrastructure in Costa Rica. It will not be difficult to sell GNP as a major
tourist attraction, since GNP is an extraordinarily beautiful place and will
become extraordinarily interesting as well. However, it will require major
improvements in roads and other minimal facilities within GNP. Additionally,
a small amount of dry forest "affirmative action" will be necessary, so that
the tourism world does not come to view Costa Rica as clothed only in
rainforest.
D. LAND ACQUISITION

The biologically correct and socially most desirable procedure would be

78

to immediately freeze all habitat perturbation ( except to leave the cattle

in certain areas) and purchase all private lands in GNP. It would then take
1-2 years to fully develop a constitution and the detailed management
technologies for GNP, except that it is obvious that the fire control program
around the park margins would begin immediately in October-November
1986. Such a plan likewise assumes that the endowment fund is in place and
functioning.
However, the human world does not function for either its own best
interest or that of the biological community it occupies; GNP acquisition
will have to proceed piecemeal as funds become available. Hacienda Oros{
is being gradually donated, but this gracious gift does not alleviate the worry
over the volcano sides, since Hacienda Oros{ contains only about 15% of the
rainforest block. All owners other than those of Santa Elena, Finca Jenny and
the Colonia on Cerro El Hacha have graciously cooperated with the GNP
plan by agreeing not to pursue active development in 1986. The owner of
Rosa Marfa has promised to attempt to avoid pesticide runoff into Santa
Rosa in 1986.
What crises are there before the end of 1986? Today, as you read this,
the tiny patches of pristine semi-evergreen forest on the sides of Cerro El
Hacha are being cut by members of the Colonia who are prepanng new
fields for corn, rice and beans. The owner of Finca Jenny could decide at
any moment to convert her forest to sawlogs or cashew plantations. The
rentors of Rosa Marfa's croplands may not wish to abide by the owner's
restrictions on pesticide use . A forest guard must be hired and provisioned
to patrol the donated portions of Hacienda OrosL We have no promises or
understanding from the owners of Santa Elena.
The Cerro El Hacha situation must be placed at the toe of the emergency
list. Barring unforseen events, Finca Jenny and the few ha of Rosa Maria
should be next. Poca Sol, Centeno and San Josecito-Tempisquito should
follow. El Hacha will hopefully be donated and major portions of Santa
Elena should be last.
In addition to the above purchases, it is imperative that Islas Murcielagos
be decreed part of Murcielago National Park, and that the southern boundary
of the Orosi Forest Reserve be reinstated so as to avoid gradual invasion by
neighboring landowners in that area . The rainforested Atlantic side of Volcan
Orosi and Volcan Cacao (apparently unoccupied and of questionable ow-
nership) must be explored as a possible inclusion in GNP so as to maximize
protection of the evergreen forest on the Pacific side of the volcanos.
It is important that the large properties be purchased as single blocks.
There are small patches of valuable real estate on each of them , and many
of these patches are of extreme biological value because of their water, soil
type, pristine forest, etc . If the large property becomes broken up as a
consequence of partial sale, then these small pieces will be sold to purchasers
who will hold them as investment property at astronomical prices (if they will
sell them at all); such inholdings are to be zealously avoided. Additionally,
the willingness of several owners to sell to GNP is based on the assumption
that the entire property will be purchased.
Land values will be determined by open market values in Costa Rica,
withthe pr i<;;esestablished by 8overnment assessors.
E. BUDGET
1. LAND PURCHASE
Including the area to be donated in Hacienda Orosi, there are approxi-
 79

mately 470 km 2 to be obtained. Assuming the acquisition of Orosi (30 km 2 )

to be successful through donation, the cost for the land to be purchased to
form GNP will be $8,800,000, assuming an average figure at $200 per ha ($81
per acre). This per ha figure is representative for undeveloped low-grade wild-
lands and farmland throughout the country at the present time.

2. ENDOWMENT

Management costs for GNP will be a minimum of $300 000 per year .
This means a minimum start-up endowment of $3 000 000. It is assumed that
this endowment fund will continue to grow after the establishment of GNP
through use fees (tourists, researchers, seed bank developers), donations, cat-
tle rental fees, publication sales, etc.

3. ORGANIZATION AND FUNDING CAMPAIGN

All costs for this campaign are being borne by personal contributions,
the Nature Conservancy International Program, the National Park Foun-
dation of Costa Rica , Fundacion Neotrc:Spicade Costa Rica, and the National
Science Foundation of the US.
4. SOURCES OF FUNDS
Funds are being sought throught through public campaign presentations
and application to foundations, individual donors and governmental institu-
tions throughout the world. Contributions are tax-deductible in the US and
may be sent to "Nature Conservancy Guanacaste Fund, 1785 Massachusetts
Ave., NW, Washington, D.C. 20036", or to "Parque Nacional Guanacaste,
Fundacion Neotropica, Apdo. 236, San Jose 1002, Costa Rica".

5. COSTA RICAN-StJPPORT

All possible connections between Guanacaste National Park and the

many relevant sectors of Costa Rican society are being actively explored at
the present time. While these connections are not likely to result in direct
financial support of GNP, they will be a critical part of the social approval and
local indirect support that are essential for GNP establishment and survival.

F. ENDORSEMENTS

The GNP development plan has been discussed and described widely
within Costa Rica and the plan as here presented incorporates feedback . No
governmental or private opposition has been identified. Supporting letters
from the Costa Rican National Park Service, the National Park Foundation of
Costa Rica, and Fundacion Netropica are attached (Appendix 2). Within the
US, the Nature Conservancy International Program is the official administra-
tor of the project, while in Costa Rica the same role is played by the National
Park Foundation and Fundacidn Neotropica.
Until substantial funding is in hand, it is inappropriate to ask for final
approval and direct involvement from the populace of the GNP region, since
GNP cannot pay its own way at present. On the other hand, during 1986 a
number of the educational aspects of the GNP plan will be developed on a
trial basis using Santa Rosa National Park, its personnel and its researchers as a
resource base,
 81

CONTINGENCYPLANS

1) What if GNP cannot be obtained? We retreat to Santa Rosa (MurcieJago

will be roasted off the map by the wildfires from Santa Elena) and carry
out all of the philosophical and educational goals for GNP on an inferior
scale and in a gradually decomposing habitat. All of the inventory and
other biological studies for GNP will still be priceless as salvage biology,
and at least tell future generations what they lost.
2) What if the land can be acquired but endowment funds cannot be located?
We use a skeleton staff to keep the fires out of GNP, rely on aerial photo-
graphy for baseline reference, and maximize the educational effort. We
grow cows for meat and management, and get out there and find the
endowment funds; the progressive agriculturization of Guanacaste and
all of Costa Rica acts in our favor in this case; when the national and
international audience can see only clean croplands except in GNP,
willingness to pay to maintain GNP will increase. Likewise, the general
educational level in Costa Rica, and the international sophistication of
the international users of tropical biology is steadily growing.
3) What if only sufficient funds for some land are available? We follow the
purchase priorities established earlier under Land Acquisition. With
Cerro El Hacha we save both the unique vegetation of the virgin semi-
evergreen forests and save a major dry season insect refugium. With the
southern margins of Santa Rosa secure, we avoid further agrochemical
contamination and save a very important canyon forest. With the Poco
Sol/Centeno/San Josecito-Tempisquito block, we save the everflowing
revers and the major transition zone between the evergreen forest on the
sides of the volcanos and the dry forest that covers the bulk of GNP.
With Santa Elena, we allow forest restoration to start on both Santa
Elena and Murcielago, save numerous very dry unique hillside and
ocean-edge habitats and protect the northern boundary of Santa Rosa.
4) What if we have a run of drought years and then a colossal fire gets away
from us? So, what's new? GNP has already been roasted hundreds of
times before. We lose some ground, but a stump that has had five years
without a fire has a healthy root system. Furthermore, all the explicitly
experimental areas can be kept free from fire by backfiring once the
battle is obviously lost. Finally, every year that passes without a fire, the
forest advances more and the overall grass area is reduced.
5) What if the poachers are undefeatable? In the short run we lose some
animals but we do not lose the breeding population (the same applies to
the marineturtleeggs).In the longtermthereis no reasonwhypoacher
82

intrusions cannot be lowered to the level found in national parks around

the "developed" world.
6) What if serious squatter pressure develops? Squatters have never been a
problem in Costa Rica on government or private land under conspicuous ·
use. The GNP managerial staff will be more than mere employees, they
will be part owners in a very real sense. Furthermore, all indications are
that in the social game of property invasion, the populace in the area of
GNP will be largely on the side of the park. In the worst scenario, GNP
might lose some marginal land to squatters. However, Costa Rica has
already lost almost all of its dry forest to agriculturization. While restora-
tion of any of this land to forested wildlands, there is nowhere to go but
up.
7) What if GNP is not big enough? Then we lose some species. So be it The
world will not give Guanacaste Province back to nature. However, no
known species in the area will be lost.
8) What if one of the volcanos erupts? It will be the first time in history
that a tropical volcanic eruption has been laid down on a documented
wild landscape.
9) What if the conflicts to the north spread into Costa Rica? The history
of Mesoamerica suggests that care of GNP might be delayed and reduced,
but if there are concerned persons in the area, minimal maintenance can
be continued during the conflict. Furthermore, an extremely effective
barrier to social strife is a resident population that knows itself well and
is satisfied with its resource base. The cultural opportunities offered by
GNP are part of bringing a population to this stage.
10) What if the government of Costa Rica should change its overall emphasis
on enlightened development and turn against a project such as GNP?
Such an event is about as likely in Costa Rica as in the US. Were Costa
Rica to lose its substantial population of citizens who already view
national parks and other kinds of preserves as highly desirable parts of
the landscape, GNP would be threatened just as would he the other
national parks. However, a major activity of GNP is producing environ-
mental awareness in the next generation of Costa Rican decision-makers
and their off spring.
11) What if the biologists lose interest in GNP? GNP needs people with bio-
logical understanding and training for its management and display to the
public. With independent funding, GNP will always attract an interested
managerial staff. This in tum, plus the biological properties of GNP, will
always serve as a magnet to biologists from within as well as without the
tropics.
12) What if the director or the executive committee or the GNP constitution
fails to function properly? As part of GNP development, all three of these
components will have provisions to assist in the replacement of non-
functioning components. With the executive committee constituted of
members from Costa Rican institutions with strong mandates in the
areas in which GNP has conspicuous offerings, disinterest seems very
unlikely.
 83

PUBLISHED INFORMATION
ON GUANACASTE DRY FOREST

There is virtually no biological literature on any part of GNP except for

Santa Rosa National Park. However, this situation will change rapidly. A
bibliography of papers on the biology (including agriculture) of Guanacaste
dry forest habitats is in preparation and already contains over 500 citations
(it will be available by the end of November 1986 from D.H. Janzen, Depart-
ment of Biology, University of Pennsylvania, Philadelphia, Pa. 19104). Santa
Rosa contributes more than 200 of these references. A few are listed below.
There is also a very large amount of (as yet) unpublished information on Santa
Rosa. I am preparing a book on Santa Rosa biology for the Costa Rican open
university (UNED). The book "Costa Rican Natural History" (1983, D.H.
Janzen, editor) is being translated to Spanish and will be published in late
1986; it contains over 100 individual species accounts and discussions of the
ecology of many major groups of organisms for Guanacaste.

SELECTED BIBLIOGRAPHY

Banco Central ( de Nicaragua) 1980. Boletfo Nicaraguense de Bibliograffa y

Documentacion, 37-38. Biblioteca, Banco Central de Nicaragua, Mana-
gua, 126 pp.
Barquera, J .I. 197 5. Plan maestro para el establecimiento y manejo del area
del Volcan Masaya como parque nacional.BancoCentralde Nicaragua,
Managua, 145 pp.
Bergoeing, J.P. 1978. Penfnsula de Santa Elena, Costa Rica. Estudio geomor-
fol6gico preliminar. Informe Semestral, Instituto Geografico Nacional,
Julio a Diciembre, pp. 19-30.
Bergoeing, J.P. L.G. Brenes and E. Malavassi 1983. Geomorfolog{a del Pacffico
Norte de Costa Rica ( explicacion del mapa geomorfologico 1: I 00 000).
Oficina de Publicaciones, Universidad de Costa Rica, Ciudad Universi-
taria, 110 pp.
Bonilla, A. 1983. Reservas forestales y zonas protectoras. Editorial Tecnolo-
gica de Costa Rica, Cartago, 312 pp.
Bonoff, M.B. and D.H. Janzen 1980. Small terrestrial rodents in eleven
habitats in Santa Rosa National Park, Costa Rica. Brenesia 17: 163-
174.
Boucher, D.H. 1982. Seed predation by mammals and forest dominance by
Quercus oleoides, a tropical lowland oak. Oecologfa 49:409-414.
Boza,M.A.and R. Mendoza1981.The National Parks of Costa Rica. INCAFO,
S.A., Madrid, Spain. 310 pp.
84
Braithwaite, R.W. 1985. Biological research for national park management.
Proceedings of the Ecological Society of Australia 14 (in press).
Braithwaite, R. W. and J. A. Estbergs 1985 . Fire patterns and woody vegeta -
tion · trends in the Alligator Rivers region of northern Australia. In
Ecology and management of the world's savannas, J. C. Tothill and
J. J. Mott, eds., Australian Academy of Science, Canberra, pp. 359-
364.
Braithwaite, R. W., M.L. Dudzinski, M.G. Ridpath, and B. S. Parker 1984.
The impact of water buffalo on the monsoon forest ecosystem in Kakadu
National Park (Australia). Australian Journal of Ecology 9:309-322 . .
Chavez, E.S. 1979. Analisis comparative de cinco comunidades vegetales
del Parque Nacional Volcan Masaya. Universidad Centroamericana,
Managua, Nicaragua, 78 pp.
Corea, E. and I. Chacon 1984. Descripci6n y observaciones preliminares del
comportamiento y h~bitat de una nueva subespecie de Morpho
polyphemus (Lepidoptera: Morphinae) de Costa Rica. Brenesia: 22:183-
188.
Cornelius, S.E. 1986. The sea turtles of Santa Rosa National Park. INCAFO,
• S.A., Madrid, Spain, 64 pp.
Fedigan, L.M. 1986. Demographic trends in the Alouatta palltata and Cebus
capucinus populations of Santa Rosa National Park, Costa Rica. In
Primate Conservation, J. Else and P. Lee, eds., Cambridge University
Press, Cambridge, England (in press).
Fedigan, L.M., L. Fedigan, and C. Chapman 1986. A census of Alouatta
palliata and Cebus capucinus in Santa Rosa National Park, Costa Rica.
Brenesia: 23:309-322.
Flemin.&, T.H. 1981. Fecundity, fruiting pattern, and seed dispersal in Piper
amalago (Piperaceae), a bat-dispersed tropical shrub. Oecologfa 51:
42-46.
Fleming, i.H. 1986. Secular changes in Costa Rican rainfall: correlation
with elevation. Journal of Tropical Ecology (in press).
II, Geology, pp. 102-114.
Fleming, T.H. 1981. Frugivorous bats, seed shadows, and the structure of
tropical forests. Biotropica 13:45 -53.
Fleming, T.H. 1985. Coexistence of five sympatric Piper (Piperaceae) species
in a Costa Rican dry forest. Ecology 66:688- 700 .
Fleming, T.H. 1985. A day in the life of a Piper-eating bat. Natural History
94(6):52-56.
Glander, K., L.M. Fedigan, L. Fedigan and C. Chapman 1986. Capture and
marking of three species of primates in Costa Rica. American J oumal
of Primatology (in press).
Gudmundson, L. 1983. Hacendados. politicos v precaristas: La filJ.naderia y
el latifundismo Guanacasteco, 1800 -1950. Editorial Costa Rica, San
Jose, 255 pp.
Harrison, J.V. 1953. The geology of the Santa Elena Peninsula in Costa Rica,
Central America. Proceedings of the Seventh Pacific Science Congress,
II, Geology, pp. 102- 104
Hallwachs. W, 1986. Agoutis ( Dasyprocta punctata): the inheritors of guapinol
(Hymenaea courbaril: Leguminosae). In Frugivores and seed dispersal,
A. Estrada and T.H . Fleming, eds., Dr. W. Junk, Dordrecht , Holland,
p_p. 205 -304.
Hartshorn, G .S. 1983. Plants. In Costa Rican Natural History, D.H . Janzen,
ed., University of Chicago Press, Chicago, pp . 118-157 ..
 85

Hedstrom, I. and T. Elmqvist 1984. Prepona butterflies (Nymphalidae) and

Hoplopyga beetles (Scarabaeidae) on the same food source during the
neotropical dry season - a case of commensalism? Revista de Biologia
Tropical 32:313-316.
Janzen, D.H. 1981. Enterolobium cyclocarpum seed passage rate and survival
in h.9rses, Costa Rican Pleistocene seed dispersal agents. Ecology 62:
593-601.
Janzen, D.H. 1982a. Gu{a para la identificaci6n de mariposas nocturnas de
la familia Satumiidae del Parque. Nacional Santa Rosa, Guanacaste,
Costa Rica. Brenesia 19/20:255-299.
Janzen,D.H. 1982b.Removalof seeds fromhorsedungby tropicalrodents:
influence of habitat and amount of dung. Ecology 63: 1887-1900.
Janzen, D.H. 1982c. Differential seed survival and passage rates in cows
and horses, surrogate Pleistocene dispersal agents. Oikos 38: 150-156.
Janzen, D.H. 1983a. No park is an island: increase in interference from
outside as park size decreases. Oikos 41 :402-410.
Janzen, D.H. 1983b. Seasonal change in abundance of large nocturnal dung
beetles (Scarabaeidae) in a Costa Rican deciduous forest and adjacent
horse pasture. Oikos 41 :274-283.
Janzen, D.H., ed. 1983c. Costa Rican Natural History. University of Chicago
Press, Chicago, 816 pp.
Janzen, D.H. 1984a . Weather-related color polymorphism of Rothschildia
lebeau (.Saturniidae). Bulletin of the Entomological Society of America
30: 16-20 ..
Janzen, D.H. 1984b. Natural history of H_ylesia lineata (Saturniidae:
Hemileucinae) in Santa Rosa National Park, Costa Rica. Journal of the
Kansas Entomological Society 57:490-514 .
Janzen, D.H. 1984c. Two ways to be a tropical big moth: Santa Rosa
saturniids and sphingids. Oxford Surveys in Evolutionary Biology 1:
85-140.
Janzen, D.H. 1985a. Spondias mombin is culturally deprived in megafauna-
frne forest. J oumal of Tropical Ecology 1: 131-155.
Janzen, D.H. 1985b. A host plant is more than its chemistry. Illinois Natural
History Survey Bulletin 33:141-174.
Janzen, D.H. 1986. Mice, big mammals, and seeds: it matters who defecates
what where. In Frugivores and seed dispersal, A. Estrada and T.H.
Fleming, eds., Dr. W. Junk, Dordrecht, Holland, pp. 251-271.
Janzen, D .H. and R. Liesner 1980. Annotated check -list of plants of lowland
Guanacaste Province, Costa Rica, exclusive of grasses and non-vascular
cryptogams. Brenesia 18: 15-90.
Janzen, D.H. and P.S. Martin 1982. Neotropical anachronisms: the fruits
the gomphotheres ate. Science 215: 19-27.
Janzen, D.H. and P.G. Waterman 1984. A seasonal census of phenolics,
fibre and alkaloids in foliage of forest trees in Costa Rica: some factors
influencing their distribution and relation to host selection by Sphingidae
and Saturniidae. Biological Journal of the Linnean Society 21 :439-454.
Lemaire, C. 1982. Trois Saturniidae inedits du Costa Rica et du Perou
(Lepidoptera). Revue Francaise Entomologie (N.S.) 4:79-85.
Rodas, J. G. F. 1985. Diagn6stico del Sector Industrial Foresta!. Editorial
Universidad Estatal a Distancia, San Jose, 116 pp.
Sader, S.A. and A.T. Joyce 1984. Relationship between forest clearing and
biophysical factors in tropical environments; implications for the design
of a forest change monitoring approach. Report No. 230, Earth Resources
Laboratory, NASA, NSTL, Mississippi, 20 pp.
 86

Sheehan, W. 1984. Nesting biology of the sand wasp Stictia heros

(Hymenoptera: Sphecidae: Nyssoninae) in Costa Rica. Journal of the
Kansas Entomological Society 57 :377-386.
Stiles, F.G. 1983. Birds. In Costa Rfoan Natural History, D.H. Janzen, ed.,
University of Chicago Press, Chicago, pp. 502-530.
Taylor, J .A. and D. Tulloch 1985. Rainfall in the wet-dry tropics: extreme
events at Darwin and similarities between years during the period 1870-
1983 inclusive. Aust. J. Ecol. 10:281-295.
Teran, F. and J. Incer Barquero 1964. Geographia de Nicaragua. Bank of
Nicaragua, Managua, 266 pp.
Uhl, C. and R. Buschbacher 1985. A disturbing synergism between cattle
ranch burning practice_§and selective tree harvesting in the eastern
Amazon. Biotropica 17: 265-268.
Unwin, G.L., G.C. Stocker, and K.D. Sanderson 1985. Fire and the forest
ecotone in the Herberton Highland, north Queensland. Proc. Ecol.
Soc. Australia 13:215-224.
Vaughan, C. 1983. A report on dense forest habitat for endangered wildlife
species in Costa Rica. Escuela de Ciencias Ambientales, Universidad
Nacional, Heredia , Costa Rica. 66 pp.
Waterman, P.G. and E.N. Mahmoud 1985 .. Flavonoids from the seeds of
Lonchocarpuscostaricensis.Phytochemistry 24:571-574.
Westoby, M. 1984. Constructive ecology: how to build and repair · ecosystems .
AES Working Paper, School of Australian Environmental Studies,
Griffith University, Brisbane, 34 pp.
Wheelwright, N.T. 1983. Fruits and the ecology of resplendent quetzals.
The Auk 100:286-301 ..
Williams, K.D. 1984. The Central American tapir (Tapirus bairdii Gill),
in northwestern Costa Rica. Ph.D. Thesis, Michigan State University,
84 pp.
Witsberger, D., D. Current and E. Archer 1982. Arboles del Parque Deininger.
Ministerio de Educaci6n, San Salvador, El Salvador, 336 pp.
 87

WHAT DOESGNP MEAN

TO COSTARICA?

Properly developed, Guanacaste National Park will mean different

things to many different people. Below we list some of these things,
recognizing that the list has overlapping parts and is not exhaustive.

1) GNP will be a major regional, national and international cultural

center. For many people of the region, it will be the difference
between living as a physically healthy human draught animal living
in rich but mindless agricultural pastures, and living as a cultured
human being. Simultaneously, its successes and failures will off er
examples to others developing the biological cultural and educa-
tional potential of other tropical sites in and out of Costa Rica.
2) GNP will be the first Costa Rican national park designed from the
beginning as a cultural and educational resource. Simultaneously,
it will be a major opportunity for Costa Ricans to put their tradi-
tional respect for education to work for themselves, rather than
simply use their educations to respond to the cultural pablum
served up by the steadily homogenizing public media.
3) GNP will demonstrate that the Costa Rican government has the
foresight and flexibility to develop its national parks rather than
simply to form and patrol them by decree.
4) GNP will be the first example anywhere in the tropics where a
small and endangered habitat was given back a large area to reinvade
and thereby get its population densities back to a more resilient
level.
5) GNP will be the only dry forest reserve in Mesoamertca large
enough to maintain healthy breeding populations and normal
habitats of the animals, plants and habitats that were here when
the Spaniards arrived. Whether it eventually becomes the only
one in the Neotropics depends largely on how much of an inspira-
tion it is to other regions to attend to their dry forests before they
disappear.
6) GNP will be the only preserved intersection of two major habitat
types, and the only preserved dry forest elevational transect, in
Mesoamerica (if not in all of the Neotropics).
7) GNP will be a living gene bank for tens of thousands of species of
wild organisms, some of which are already of established commercial
value (e.g., timber trees, fuelwood trees, game animals) and many of
which will some day be of commercial value.
8) GNP will be a large and diverse example and data source for studies
88

and projects in reforestation with dry forest plants, and in the

manipulation of habitats to this end. It will also go a long way
toward destroying the myth that tropical humans cannot be in
control of their environmental destiny.
9) GNP will be the first neotropical national park with a substantial
endowment fund and therefore the ability to survive a variety of
economic perturbations and excercise some autonomy over its
management plans.
10) GNP will offer salaried and secure local employment that will
employ fewer people than if the land were colonized by subsistence
farmers but substantially improve the cultural lives of Costa Ricans
from local subsistence farmers to San Jose upper income residents.
Furthermore, the annual management budget of GNP will represent
a substantial cash flow into the local economy.
11) GNP will be an economic resource through significant development
of the conventional and educational tourism industry; participation
will range from local guide service and living accomodations to
international-level tours. The stress will be on the educational
aspects qf tourism.
12) GNP will show that the international community is willing to
recognize its financial and intellectual responsibility towards a
portion of the tropics, a portion that has enormouscollectivevalue
to the world at large.
APPENDIXES
 91

APPENDIX1. RELATIONSHIPOF GNP TO OTHERPACIFIC

MESOAMERICANDRY FOREST RESERVES.

Guanacaste National Park will be the largest link in the thin chain of
tropical dry habitat national parks, forest reserves, wildlife refuges, etc.
that stretches from tropical western Mexico to Panama on the Pacific coast
of Mesoamerica. Additionally, within Costa Rica it is by far the largest island
in a highly fractured archipelago of dry forest preserves. Except for Santa
Rosa and Murcielago, which will become part of GNP, these sites will all
make heavy use of GNP in the future as a reference point and will
simultaneously contain some habitats that can never occur in GNP.

OUTSIDE OF COST A RICA

1.- Estacion Biologica Chamela. 120 km north of Manzanillo, Jalisco,

Mexico. This 16 km 2 preserve and Biological Field Station is in lightly
disturbed Tropical Deciduous Forest (Bosque Cauducifolio Tropical) on
lightly dissected undulating terrain of sandy metamorphic rocks (50-100 m
elevation) . It has a well-developed biological research station (initiated
in the early l 970's) and is owned by the Instituto de Bio logia of the Universi-
dad Nacional Aut6noma de Mexico in Mexico City. Address: Estacion Bio-
16gica Chamela, Apartado Postal 21, San Patricio, J alisco 48980, Mexico.

2. - Parque Deininger. 5 km east of the city of La Libertad. La Libertad

El Salvador. This 7. 32 km 2 national park is lightly disturbed (though heavily
hunted in the past) tropical deciduous forest on steep slopes (7-300 m
elevation) and much like the deciduous forests in the central ·portion of
Santa Rosa National Park, Costa Rica. Deininger National Park is relatively
unstudied, except that its tree flora has been inventoried. Address: Lie.
Manuel Benitez Arias, J efe, Servicio de Parques Nacionales y Vida Silvestre,
MAG, San Salvador, El Salvador.

3.- Parque Nacional Volcan Masaya. 20 km southeast of Managua on

the asphalt highway from Managua to Granada, Nicaragua. This 43 km 2
national park is centered on two periodically active volcanos and clothed
with heavily disturbed deciduous forest remnants.

4.- Monumento Nacion.:] Sarigua. 235 km southwest of Panama City,

and on the tip of the Azuero Peninsula on the Pacific coast of Panama.
This new 60 Km national monumentis in the processof formationon
2

coastal foothills and shore (mangrove). The highly deciduous vegetation

has been badly perturbed by farming and clearing, but may recover once
 92
protected (Fundaci6n de Parques Nacionales y Medio Ambiente de Pana-
ma, personal communication).
In addition to the above, there are a few tens of km 2 of dry forest sites
in central or eastern Mexico that are either in biological preserves or being
considered for inclusion .

INSIDE COST A RICA

While the conservation picture in Costa Rica is still fluid and expanding,
dry forest habitats have been so thoroughly agriculturalized that there is
almost no pristine forest remaining, outside of extant preserves, that can be
used to increase conserved areas. Forest restoration is the only means by
which one can substantially increase the area of Costa Rica's dry forests
that are under protection .
1.- Parque Nacional Santa Rosa. (Santa Rosa Section).35 km north of
Liberia in Guanacaste Province. This 108 km 2 rectangular block stretches
from the Interamerican Highway to the Pacific Ocean (0-350 m elevation)
over _plateaus. canyons and coastal plain. Thr vegetation ranges from 2 m tall
·totally dry season deciduous forest to 30-40 m tall evergreen forest, with
successional stages of 0-400 years in age and numerous old pastures of
1-200 ha in extent. The site is under intensive study by biologists, and will
be a major source of inoculum for GNP. Santa Rosa was the first large
national park to be formed in Costa Rica ( 1972) and is firmly embedded in
the national park system. Address: Parque Nacional Santa Rosa, Apdo. 169,
Liberia, Guanacaste Province, Costa Rica, tel. 69-5598.
2.- ParqueNacionalSantaRosa. (Murcielago Section). Along the north
half of the Santa Elena Peninsula, Cuajiniquil, Guanacaste Province. This
122 km 2 section covers rocky mountains to low hills that were once covered
with deciduous dry forest but are now primarily covered with abandoned
pastureland. It is rich in mangrove and intertidal habitats, and still contains
enough small vegetation patches to reforest if allowed through exclusion
of fires. This new addition (1980) to Santa Rosa National Park has not
been investigated biologically. Once consolidated with Santa Rosa and the
intervening Santa Elena penin sula, Murci,$go will be a major dry forest
patch; by itself, it has no chance to escape from the dry season fires that
sweep the Santa Elena peninsula. Address: Same as Parque Nacional Santa
Rosa.
3. - Parque Nacional Palo Verde. On the flood plain and east bank of
the Ri'o Tempisque as it spreads into the Gulf of Nicoya. This 94 km 2 park
is the southern portion of the combined floodplain preserve of Palo Verde
National Park and Refugio Nacional de Fauna Silvestre Dr. Rafael Lucas
Rodnguez Caballero; this preserve was established largely to protect waterfowl.
The site is largely cleared of forest, and has been heavily grazed and burned,
but it may eventually return to approximately pristine vegetation if the fires
are halted.
4.- Refugio Nacional de Fauna Silvestre Dr. Rafael Lucas Rodnguez
CabaUero. Upriver and bordering Parque Nacional Palo Verde . This 74 km 2
wildlife refoge has great potential for preserving dry season waterfowl habitat
an the -rich flood plain flora. Its aquatic habitats are, however, severely
threatened by the agrochemical runoff from Guanacaste agriculture and by
water control in the Rfo Tempisque . Both the park and the refuge are under-
going massive vegetation changes at present, owing to removal of cattle and
intensification of the fire regime.
5.- Parque Nacional Barra Honda. 23 km NE of Nicoya in the upper
 93

Nicoya Peninsula. This 23 km 2 national park was established on severely

perturbed dry forest on limestone hills to protect an extensive cave system.
The site is biologically unstudied, but may be found to be ecologically
important for its limestone-based vegetation as well as its cave biology (and
archaeology).
6.- Reserva Natural Absoluta Cabo Blanco. The tip of the Nicoya
Peninsula (Puntarenas Province). This 12 km2 reserve was established to
protect seabird roosting and nesting sites. The forest behind the beach is
significant in a protective sense, but may also be of relictual value as well.
7. - Refugio Nacional de Fauna Silvestre de Ostional. 35 km SW of
Nicoya on the Pacific coast. This 0.16 km 2 wildlife refuge was established
for the protection of Playa Ostional's sea turtle nesting sites.
8.- Lomas Barbudal Reserva Biol6_gica. 15 km SW of Bagaces. This 30
km 2 newly established biological reserve is in a relictual deciduous forest
on low hills. While the site has been severely perturbed by hunting and logging,
it contains sufficient population relicts to eventually return to incomplete but
superficially intact dry forest if the fires are halted.
 95

APPENDIX2. OFFICIALLETTERSOF SUPPORTFOR GNP.

MINISTERIO DE AGRICULTURA Y GANADERIA

SERVICIODE PARQUESNACIONALES

S.P.N. 259
January 28, 1986

Dr. Daniel Janzen

Department of Biology
University of Pennsylvania
U. S. A.

Dear Dr. Janzen :

I enjoyed the presentation you gave on the proposed creation of Guanacaste National Park.

As discussed, the new park coincides roughly with the area recommended by a study conducted by the
Tropical Science Center on potential areas and additions to the system of national parks and reserves, a
few years ago.

The National Park Service approves and supports this project. It strives to preserve an excellent example
of TropicalDry forest and its remarkablebiologicaldiVersity, However,I want to stressthe need to
establish an endowment fund to ensure proper management and consolidation, before the area is turned
over to the Park Service.

Thank you Dan. We, and Costa Rica in general, are fortunate to have you working with us. Your
contribution to the preservation of our renewable natural resources is invaluable .

I look forward with enthusiasm to the successful outcome of this challenging endeavor and I encourage
you to keep working at it.

With kind personal regards,

Alvaro F. Ugalde
Director
96

FUNDAClON DE PARQUES NACIONALES

January 28, 1986

Dr. Daniel Janzen

Department
of Biology
University of Pennsylvania
Philadelpltia, Pennsylvania 19104
u.s. A.

Dear Dr. Janzen:

The · National Parks Foundation has reviewed your proposal for the establishment and the
subsequent management of Guanacaste National Park.

We agree that the park will serve well to preserve the last remnants of the tropical dry forest in
Latin America, the protection of this area will be an unprecedented endeavor in Latin America, and of
~eat benefit to Costa Rica and visitors from around the world.

We are eager and willing to work with you as in the past, and assure you oux full cooperation
and support.

Due to the current financial difficulties the government of Costa Rica is experiencing in
administering the system of protected areas, we recommend that your campaign include funds for land
acquisition and to set up a management fund for the park.

We thank you for your continued interest in conservation, and wish you much success with the
proposed project
Sincerely,

Mario A. Boza, President

 97

No. 1/86
January 27, 1986

Faculty of Arts and Sciences

Department of Biology
University of Pennsylvania
c/o Dr. Daniel Janzen:

Gentlemen:

We make reference to the project involving the creation and management of Guanacaste National Park.
The proposed _park includes the existing Santa Rosa National Park, Murcielago, Cacao and Orosi
Volcanos and surrounding areas. Your proposal coincides with our interest in preserving the
ecosystems already represented in Santa Rosa, but more importantly, we are concerned with the protec -
tion of those ecosyst~s found outside the current system of natural areas. In particular, we are
concerned with the tropical dry forest which is very threatened and is underrepresented, perhaps three
areas exist in all of Tropical America, that can be preserved.

We trust in success for which in advance we would like to express our most sincere appreciation.

Sincerely ,

Dr. Rodrigo Gamez

President
Fundaci6n Neotr6pica
 99

APPENDIX 3. MONTHLY PRECIPITATION (ROUNDED TO THE

NEAREST MM) IN THE ADMINISTRATION AREA OF
SANTA ROSA NATIONAL PARK, GUANACASTE PROV-
INCE, COSTA RICA (DATA COLLECTEDBY PARK
RANGERS AND EXTRACTED FROM THE METEOROLOGY
INSTITUTE IN SAN JOSE).

Year Jan. Feb . Mar. Apr. May. Jun . Jul. Aug. Sep. Oct. Nov. Dec. Tot.

1980 1 0 5 0 184 175 139 159 331 417 240 9 1660

1981 0 1 1 11 353 582 172 478 195 268 153 27 2241

1982 16 2 0 41 919 129 117 34 328 197 37 1 1820

1983 2 0 22 4 21 180 106 107 188 201 79 7 917

1984 6 8 0 0 118 218 278 162 613 261 52 7 1723

 101

APPENDIX 4. PROTECTION STATUS OF THE AREA TO BE INCLUDED

WITHIN GUANACASTE NATIONAL PARK (SOURCE:
COSTA RICAN NATIONAL HERITAGE PROGRAM).

1986

KM2 HA

Areaprotectedin NationalParks 229.90 22990 32.72%

Area semi-protected in Forest Reserves 105.45 10545 15.02%

Total protected 335.35 33535 47 .74%

Continental area not protected 363.80 36380 51.79%

Island area not protected 3.30 330 0.47%

Total unprotected 367.10 36710 52.26%

Total area of Guanacaste National Park 702.45 70245

 103

TABLEOF CONTENTS

COVER . ..... . .......... . ... .. ........ . .... .. ... .. .. . 7

SUMMARY ............... .. . . ... . ..... · · ·. . ... . ...... 9
INfRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
THE REGION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
In general . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Ecological placement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Habitats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
GNP size. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
Fauna . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
HUMAN OCCUPATION OF GUANACASTE NATIONAL PARK. . ... . . . . 31
Prehistoric . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Contemporary ownership . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Human resources in the area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
THE ACTION PLAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
A. Allow forest reinvasion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
1. Fire . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
2. Seed movement ~indseedling establishment . . . . . . . . . . . . . . . 72
3. Intensively managed reforestation . . . . . . . . . . . . . . . . . . . . . 73
4. Cessation of hunting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
B. Organizationalinfrastructure. . . . . . . . . . . . . . . . . . . . . . . . . . 74
C. Use programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
1. Information Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
2. Inventory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3. Research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4. Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
J. UseZoning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
6. Educational Facilities. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
D. Land acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
E. Budget . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Land Purchase . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1. 78
2. Endowment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
3. Organization and Funding Campaign . . . . . . . . . . . . . . . . . . 79
4. Sources of Funds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
5. CostaRican Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
F. Endorsements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
CONTINGENCYPLANS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
PUBLISHEDINFORMATIONON GUANACASTEDRY FOREST. . . . . . . . . 83
Selected bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
WHAT DOES GNP MEAN TO COSTA RICA . . . . . . . . . . . . . . . . . . . . . . 87
Appendix 1. Other dry forest reseroesin Mesoamerica. . . . . . . . . . . . . . . . . . en
Appendix 2. Official letters of support . .. , .... , ... , ... , , . , . . . . . . . 95
Appendix 3. Santa Rosa National Parkprecipitation . . . . . . . . . . . . . . . . . . . 99
Appendix 4. Areal extent of GuanacasteNational Park . . . . . . . . . . . . . . . . . 101
Este libro se termin6 de imprimir en el mes de octubre de 1986 en los Talleres Graficos
de la Editorial EUNED. Su edici6n consta de 4000 ejemplares, imprcsos en papd bond
75 gramos con forro de cartulina lino. Estuvo al cuidado de la Direcci6n Editorial de la
UNED.
El material gr2.fico pcnenecc al au tot.
Los dibujos de los mapas son rcproducciones rcalizadas por Bernal Murillo.
 ACI
ON
)
C TINKER

FUNDACION TINKER
Nueva York, E .U.A.

I
UNIVERSIDAD ESTATAL A DISTANCIA
Programa de Educaci6n Ambiental

FUNDACION DE PARQUES NACIONALES

I
EDITORIALUNIVERSIDAD
ESTATALA DISTANCIA