Content uploaded by Miguel Angel Martínez Morales
Author content
All content in this area was uploaded by Miguel Angel Martínez Morales on Feb 28, 2015
Content may be subject to copyright.
THE COZUMEL CURASSOW:
ABUNDANCE, HABITAT PREFERENCE
AND CONSERVATION
BY
MIGUEL ANGEL MARTINEZ-MORALES
A dissertation submitted to the University of Cambridge, in partial fulfilment of the
conditions of application for the Degree of Master of Philosophy.
August, 1996
Wildlife Research Group Wolfson College
Department of Anatomy Cambridge
University of Cambridge
ii
CONTENTS
Page
Preface vi
Acknowledgements vii
SUMMARY 1
1 INTRODUCTION 2
1.1 The Cracids 2
1.2 Distribution and habitat of Crax rubra 3
1.3 Conservation status of Crax rubra 3
1.4 Taxonomic status of Crax rubra griscomi 4
1.5 Research justification and aims 5
2 COZUMEL ISLAND 7
2.1 Geological history 7
2.2 Hydrology 9
2.3 Climate 9
2.3.1 Temperature 9
2.3.2 Rainfall 10
2.4 Hurricanes 11
2.5 Vegetation 12
2.5.1 Tropical semi-deciduous forest 13
2.5.2 Low tropical deciduous forest 13
2.5.3 Mangrove forest 14
2.5.4 Other vegetation types 14
2.5.5 Secondary vegetation 15
2.6 Endemic vertebrate fauna 15
iii
2.7 Human background on the island 17
3 METHODS 20
3.1 Field data collection 20
3.2 Adverse factors data collection 22
3.3 Data analysis 23
3.3.1 Density and population size estimation 23
3.3.2 Assessment of cenotes influence on the Cozumel curassow
population 24
3.3.3 Assessment of human influence on the Cozumel curassow
population 25
3.3.4 Assessment of vegetation influence on the Cozumel curassow
population 25
4 RESULTS 26
4.1 Density of the Cozumel curassow 26
4.2 Population size of the Cozumel curassow 28
4.3 Cenotes influence on the Cozumel curassow population 30
4.4 Human influence on the Cozumel curassow population 32
4.5 Vegetation influence on the Cozumel curassow distribution 35
4.6 Natural factors affecting the Cozumel curassow population 36
4.6.1 Hurricanes 36
4.6.2 Wildfires 37
4.7 Anthropogenic adverse factors 38
4.7.1 Hunting 38
4.7.2 Habitat disturbance 39
4.7.3 Exotic fauna 40
5 DISCUSSION 42
5.1 Cozumel curassow density and other curassow densities 42
5.2 Population size of the Cozumel curassow 44
iv
5.3 Habitat preference 45
5.3.1 Cenotes influence on the Cozumel curassow abundance and
distribution 45
5.3.2 Human influence on the Cozumel curassow abundance and
distribution 46
5.4 Hurricanes and wildfires 46
5.5 Hunting 48
5.6 Present habitat disturbance 49
5.7 The boa an introduced predator 50
5.8 Conservation status of the Cozumel curassow 50
6 CONSERVATION OF THE COZUMEL CURASSOW 52
6.1 Recommended actions for the conservation of the Cozumel curassow 52
6.1.1 Hunting banning enforcement 52
6.1.2 Natural protected areas and habitat enhancement 53
6.1.3 Eradication of exotic fauna 53
6.1.4 Environmental education 54
6.1.5 Captive breeding and reintroduction programme of the Cozumel
curassow 54
6.2 Research needs 55
LITERATURE CITED 57
APPENDIX I
Morphometric and plumage comparisons of the Cozumel curassow and the great curassow
APPENDIX II
Life and fertility tables of the Cozumel curassow
v
APPENDIX III
Cenotes influence on the Cozumel curassow density
APPENDIX IV
Human influence on the Cozumel curassow density
vi
PREFACE
I certify that this Master of Philosophy Thesis was conducted from the Wildlife Research
Group, Department of Anatomy, University of Cambridge, under the supervision of Dr. David
J. Chivers.
I hereby declare that my thesis is the result of my own original work, except where
explicitly stated otherwise. This thesis is not substantially the same as any that I have submitted
for a degree or diploma or other qualification at any other University. I further state that no part
of my thesis has already been or is being concurrently submitted for any such degree, diploma
or other qualification.
I also declare that my thesis does not exceed 15,000 words, the limit of length prescribed
in the Special Regulations of the M. Phil. examination for which I am a candidate.
vii
ACKNOWLEDGEMENTS
The field work of this study was possible thanks to the financial and logistical support
provided by the Instituto de Biología, Universidad Nacional Autónoma de México; CONABIO,
National Fish and Wildlife Foundation, United Nations Environmental Programme (Caribbean
Unit), Idea Wild, and Benotto.
I am particularly indebted to Dr. Patricia Escalante for her support and friendship, as well
as to Tania Macouzet and Adriana Amador from the Instituto de Biología, UNAM.
In Cozumel Island, logistical support was kindly provided by the Municipal Government,
the Municipal Police, the Comisión de Agua Potable y Alcantarillado (CAPA) and the
Fideicomiso Caleta de Xel-Ha y del Caribe (FIDECARIBE). Special thanks are due to the
Fundación de Parques y Museos de Cozumel, particularly to the personnel of the Museo de la Isla
de Cozumel, for their logistical support during the field work, to facilitate the spread of the
information obtained in this study among the inhabitants of Cozumel Island, and their friendship.
Valuable advise to work in the island's forests was received from Ramón Poot, Don Inés
Cárdenas and many other local informants, to whom I am grateful.
My studies at the University of Cambridge were financially supported by the Consejo
Nacional de Ciencia y Tecnología from México, and the British Chevening Scholarship from The
British Council.
This thesis was supervised by Dr. David J. Chivers from the University of Cambridge,
to whom I am indebted for giving me the opportunity of being a member of the Wildlife Research
Group. I am also grateful to Alfredo D. Cuarón, as my national adviser from the Universidad
Nacional Autónoma de México, who spent long hours of his time helping me with the analyses
of my results and reviewing the manuscripts. Mauro Galetti, read earlier drafts of my thesis and
contributed with useful comments. I also appreciate the friendship of all members of the Wildlife
Research Group, with whom I shared nice times during my desk work.
The Natural History Museum (British Museum), the American Museum of Natural
History and the National Museum of Natural History (Smithsonian Institution), kindly let me
review their cracid specimens and use their facilities. The visits to the museums were founded
by the Sutasoma Trust (Palmer Wheeldon Solicitors) and the Frank M. Chapman Memorial
Found (American Museum of Natural History).
TO THE CEIBA
1
SUMMARY
The Cozumel curassow (Crax rubra griscomi) a subspecies endemic to Cozumel Island,
was considered as possibly extinct since the late 1940s, because of the lack of records. However,
the present extent of the island's forested area (about 75% of 486 km ) and the occasional records
2
of this curassow by local people and researchers visiting the island, contradicted this belief.
In order to estimate the density and population size of C. r. griscomi on Cozumel Island,
I carried out a study in the tropical semi-deciduous and low tropical deciduous forests of
Cozumel Island, by surveying 386 km along line-transects, from October 1994 to June 1995. The
perpendicular distance to the line-transect, and the exact position of the bird along the transect,
were recorded to estimate density, and to relate its location with habitat characteristics,
respectively. Density estimation was based on the distance sampling theory, and the programme
DISTANCE V2.2 was used for the calculations. Population size was estimated from the density
obtained and the extent of the island's forests (349 km ). Additionally, information about natural
2
and anthropogenic adverse factors was gathered to find any relation between them and the present
population status of the Cozumel curassow.
The estimated density was 0.87 curassows/km and the corresponding population size of
2
about 300 (164-562) birds (about 137-468 adult birds). These numbers and the fact that the
Cozumel curassow is restricted to the island, place this cracid within the conservation status of
Critically Endangered, and show that this population is one of the smallest within the cracids.
This curassow showed a preference for the tropical semi-deciduous forest of the island.
Freshwater sources within the forests are showed to have a positive influence on the Cozumel
curassow distribution within a radius of 2 km. Likewise, the older stands of forest and
specifically those where Manilkara zapota and Mastichodendron foetidissimum are particularly
abundant, have a positive influence on C. r. griscomi distribution. On the other hand, human
settlements and human accesses to the forest interior showed a negative influence on the
distribution of this cracid within approximately 4.5 km.
Hunting has been probably the most important cause of reduction in the Cozumel
curassow population and, thus, is likely to be the responsible of its present conservation status.
Recommendations for the conservation of the Cozumel curassow and its habitat, are
presented. This cracid is proposed as a case study of insular biota.
2
1
INTRODUCTION
1.1 THE CRACIDS
The family Cracidae (curassows, guans and chachalacas) is an endemic Neotropical group
of large, forest-dwelling, frugivorous birds. The family is distributed from southern Texas,
U.S.A., through Central and South America to northern Uruguay and Argentina.
This group is composed of 50 recognized species (Monroe and Sibley, 1993; Hoyo et al.,
1994; Strahl et al., 1995) assembled in 11 genera: Ortalis, Penelope, Pipile, Aburria,
Chamaepetes, Penelopina, Oreophasis, Nothocrax, Mitu, Pauxi and Crax. Cracids are among
the most endangered avian families in the Neotropics, since 1 species is extinct in wild, 7 species
are currently listed as critically endangered, 5 species are considered as endangered and 15
species are defined as vulnerable (Strahl, 1990; Strahl and Grajal, 1991; Collar et al., 1992;
Ferguson-Lee and Faull, 1992; Sick, 1993; Collar et al., 1994; Strahl et al., 1995). It means that
56% of the family species are classed within a category of conservation concern.
The main two factors that have contributed heavily to the rapid decline of cracids over
the past several decades are hunting and habitat destruction (sometimes closely related to human
population growth). Populations of cracids in nature have consistently shown the inability to
maintain themselves under continuous hunting (especially in combination with habitat
destruction), because of their low reproductive rate (Delacour and Amadon, 1973; Silva and
Strahl, 1991). Furthermore, in practically all studies of hunting in Neotropical forests, cracids
suffer the heaviest pressure in the avian community and in some places this has led to local
extinctions (Redford and Robinson, 1987; Strahl, 1990; Silva and Strahl, 1991; Redford, 1992).
On the other hand, since ecological requirements of most cracids rely mainly on primary forests,
these birds are particularly susceptible to habitat destruction and fragmentation.
3
1.2 DISTRIBUTION AND HABITAT OF Crax rubra
The range of Crax rubra, the great curassow, is from eastern México, south through
Central America to the extreme north west of South America. It reaches from approximately
latitude 24/ N in the region north-west of Ciudad Victoria in southern Tamaulipas, México, south
along the forests of the Sierra Madre Oriental of eastern México, to the east of the central plateau
of the Isthmus of Tehuantepec, thence south through Central America, and northern South
America west of the Andes in Colombia (eastward only to longitude 76/ 71' W in the valley of
the upper Río Sinú) and probably to northwestern Ecuador (formerly to Guayas and the Chongon
Hills of western Ecuador). In México the range includes the Yucatán Peninsula and Cozumel
Island (Vaurie, 1967; Delacour and Amadon, 1973; Blake, 1977; Strahl et al., 1995).
C. rubra, inhabits mature, undisturbed tropical forests and mangrove forests, in tropical
and subtropical zones. It is found from sea level up to about 1,500 metres and sometimes higher,
e.g., in Volcán Barú, Panamá, the species has been recorded up to 1900 metres. In the Yucatán
Peninsula, Cozumel Island and parts of Costa Rica also occurs in seasonally drier forests.
Occasionally ventures into ravines, and, if unmolested, into partially cleared areas or even
plantations (Delacour and Amadon, 1973; Blake, 1977; Hoyo et al., 1994; Howell and Webb,
1995).
Two subspecies are recognised: C. r. rubra, the great curassow, and C. r. griscomi, the
Cozumel curassow. Both subspecies differ from each other in size, C. r. griscomi being smaller
(Nelson, 1926; Ridgway and Friedmann, 1946; Strahl, 1990; personal observations). C. r. rubra
is distributed in the whole range of the species, except for Cozumel Island where C. r. griscomi
occurs.
1.3 CONSERVATION STATUS OF Crax rubra
C. r. rubra is considered as an endangered or vulnerable taxa (SEDESOL, 1994; Strahl
et al., 1995). In México, it has been extirpated from much of its original range (Howell and
Webb, 1995), although in the forests of Chimalapas-Uxpanapa-El Ocote (Oaxaca, Veracruz and
Chiapas, respectively) viable populations may still exist, considering the extent of habitat
4
availability. Likewise, in the area of the Selva Maya, comprising the Selva Lacandona (Chiapas),
Calakmul (Campeche), Sian Ka'an (Quintana Roo) and the adjacent forested areas in southeastern
México; El Petén (Guatemala) and Belize, the habitat availability is of considerable extent
(Cuarón, 1991; Cuarón, personal communication) to support viable populations. In the Pacific
slope of Guatemala, the populations have been fragmented and reduced, particularly in the
Atitlán complex and the southern slopes of Lacandón and Chiquibal volcanoes. In El Salvador,
the subspecies survives in El Imposible, and in Honduras populations have been greatly reduced.
The Mosquitia region of Honduras and Nicaragua is likely to support stable populations of C. r.
rubra. In Costa Rica, it exists only in protected areas, including Santa Rosa, Rincón de la Vieja
and Corcovado. In Panamá, is quite widespread on the Caribbean slope, although on the Pacific
slope it seems to be limited to southern Veraguas, western Azuero Peninsula and eastern Darien.
It has apparently been extirpated from the Chiriquí region. In Colombia persists only in remote
areas (e.g., Chocó). In Ecuador, this cracid has been extirpated in most of its original range, and
the remaining population is being severely reduced (Hoyo et al., 1994; Strahl et al., 1995; Matola
et al., in press; Cuarón, personal communication).
The population status of the C. r. griscomi subspecies will be presented in detail in this
study.
1.4 TAXONOMIC STATUS OF Crax rubra griscomi
The origin of C. r. griscomi in Cozumel Island is still unknown. The most ancient
evidence of the occurrence of C. rubra on the island is the finding of 28 bones (plus 6 bones only
determined to genus) by Hamblin and Rea (1979) and Hamblin (1984), during archaeological
excavations in the island. They do not report an exact dating for these bones, although all the
bird bones excavated were dated between AD 300 and mid-1500s (i.e., during the Maya period
at Cozumel Island).
The presence of wild curassows in Cozumel Island was first reported by Ridgway (1885),
when Buteogallus anthracinus, the common black hawk, was observed feeding on a female
curassow. A few wing feathers of this curassow were collected, and Ridgway stated that they
were different from the corresponding ones of the mainland great curassow. Later on, also in
5
1885, three females were collected by Salvin and Gaumer (Salvin, 1889), but no differences were
found between the curassows from Cozumel Island and the mainland, being considered the same
species.
In 1901, six further specimens (three males and three females) of the Cozumel Island
population were collected by Nelson and Goldman, but it was not until 1926 (Nelson, 1926),
after the review of these specimens, that C. rubra was split into the mainland subspecies and the
Cozumel Island subspecies. Nelson considered the smaller size and some plumage features as
subspecific characteristics. He also considered the smaller size of the knob in males as a
subspecific feature, although now it is known that it is an age-related characteristic (personal
observation, but see Buchholz, 1991). Since then, the Cozumel curassow has been considered
as a valid subspecies. In recent years, however, the validity of this subspecies has been
questioned (Strahl et al., 1995), basically because of the great variability in female plumage
within the species and the lack of morphological differences between males of both subspecies.
Because the taxonomic status of the Cozumel curassow has significant implications for
the establishment of conservation priorities for its population, it became of fundamental
importance to define the validity of the C. r. griscomi subspecies. I reviewed 111 museum
specimens of C. r. rubra and the available 9 specimens of C. r. griscomi to search for
quantitative and qualitative differences between both subspecies. The differences found in some
of the characteristics analysed were large enough to validate the separation of the Cozumel
curassow as a true subspecies (Appendix I).
1.5 RESEARCH JUSTIFICATION AND AIMS
Nelson (1926) was the first to talk about the relative abundance of the Cozumel curassow
when he stated that "this curassow was common", when he and Goldman collected some
specimens of this bird in 1901. This seems to be likely, since they were able to collect six
curassows in five days, according to the collection dates of the specimens. By the late 1940s,
however, this cracid was considered almost extinct (Paynter, 1955), and by the late 1960s as a
possibly extinct subspecies (Delacour and Amadon, 1973), because of the lack of records.
6
For almost fifty years, the existence of the Cozumel curassow population was questioned
(Hoyo et al., 1994; Howell and Webb, 1995), but nobody actually did a study to assess its
population status. The fact that this curassow was sometimes hunted by local people (Paynter,
1955; Delacour and Amadon, 1973; local informants, personal communication), and the record
of a male curassow in the mangrove forest around the Laguna Ciega, in the north of Cozumel
Island (Miranda et al., 1988), suggested that the lack of records of the Cozumel curassow was,
in fact, the result of both the diminution of its population size and the lack of proper studies on
its conservation status.
If a management plan for the conservation of C. r. griscomi is to be proposed, it is
fundamental to carry out studies to monitor its population and to correlate its changes with
environmental factors that may affect the population of this curassow. This information could
also be useful for a better understanding of the dynamics of insular populations, isolated
populations, small populations; the effects of exploitation, exotic fauna and natural disasters.
Likewise, it is essential to know the habitat preferences of this cracid on the island, to
protect or enhance those habitat characteristics that might favour its population. The protection
or enhancement of suitable habitat for the Cozumel curassow may benefit other wildlife forms
as well, thus C. r. griscomi can be considered as a flagship taxon for the conservation of other
wildlife populations and their habitats on Cozumel Island.
The Cozumel curassow is a case study that summarizes many of the conservation
concerns and challenges of insular biotas. In this study, I present the results of the first
comprehensive study of C. r. griscomi density and population size and structure on Cozumel
Island, and determine objectively the conservation status of this cracid. I also relate different
density estimates and abundances with habitat characteristics, to show habitat preferences of the
Cozumel curassow on the island. I present an account of the adverse factors that very likely had
or have had some influence on the population and conservation of C. r. griscomi.
7
2
COZUMEL ISLAND
Cozumel Island (Figure 2.1) in the state of Quintana Roo, México, is a coralline
limestone block of 486 km (not including inland seasonal lagoons). The island is located 17.5
2
km off the northeastern coast of the Yucatán Peninsula, on the Caribbean Sea (20/16' to 20/36'
N and 86/44' to 87/02' W).
2.1 GEOLOGICAL HISTORY
Cozumel Island was formed by a gradual accumulation of calcareous sediments originated
from a coralline limestone block that rose as sea level increased and eroded as sea level dropped.
These calcareous sediments belong to the stratigraphic group named Carrillo Puerto Formation,
dated from the Upper Miocene to Pliocene (Butterlin and Bonet, 1963; Lesser-Jones et al., 1977).
It is likely that Cozumel Island is of Pliocene age, but there is evidence that the island was
submerged during the Pleistocene (Richards, 1937).
The island is separated from the mainland by the Cozumel Channel which is
approximately 450 metres deep. This fact apparently eliminates the possibility of an earlier land
bridge with the mainland and might justify the classification of the island as "oceanic", as has
been proposed by Davidson (1975).
The geomorphological landscape of Cozumel Island, as that of the Yucatán Peninsula,
is derived from its marine origin. The karst landscape has an early stage of erosion within the
geomorphological cycle. The erosive effect of water in the calcareous soil creates vertical or
horizontal dissolution conducts of different shape and size, which are the responsible of the
typical karst landscape. The vertical conducts exposing the underground water level are called
cenotes, a typical feature of the landscape in the Yucatán Peninsula, and the fall of horizontal
cave ceilings, produce stony depressions known as dolinas. In Cozumel Island cenotes and
dolinas are small, because the karst landscape has not reached the development level of the
mainland.
8
·S;)!1!UOlUlUO::J
SB~JBuBqJD
SB~JBP~LWOJSUBJ.L
UOI¡B¡~¡¡~A~una
UOI¡B¡~¡¡~AdWBA\S
¡S~JoJ ~A01¡¡UBW
¡S~.l0J S11011pIJ~p¡BJIdOl¡ A\0'1
¡S~JOJ S110npIJ~p-llll~S ¡BJldOJ.L
•
~
O
11
O
r.'l
1::;.:.;.:]
81
v::rs NV::r88RIVJ
,OOoL8
N
V3S NV3881<IVJ
NV3JO JIdIJVd
o
PUC[sJ
~OJ',
OJIX3~
10
1111D
W)/
O
¡
C;
O
,
".
,OE oOZ
OJIX3~
\
\
9
2.2 HYDROLOGY
There are no rivers in Cozumel Island, drainage on the island is almost completely
underground, as is common in the Yucatán Peninsula. The high water permeability of calcareous
rocks has led to the formation of an underground freshwater layer that lies on a salt water layer.
The existence of underground currents is found at La Caleta, Chankanaab and Chen Río, where
huge quantities of freshwater enter the sea from underground channels (Davidson, 1975).
Surface water can be seen after a heavy rain, but it rapidly evaporates and seeps through
the permeable limestone rock. There are a few aguadas or pot holes that retain rainwater for a
time, but there is practically no runoff.
Cenotes are the most important source of surface water for wildlife (and formerly for
human population) on the island, since they provide water all along the year, being particularly
important during the dry season. The classic form of cenote is a deep, cylindrical-shaped
sinkhole that has a ground water fed pool at the bottom, but they are also found as a grotto-like
formation featuring water beneath an overhang.
2.3 CLIMATE
Cozumel Island is influenced by warm marine waters and consistent easterly trade winds.
These two features, in addition to the small size of the island, produce rather homogeneous
climatic conditions. The island has a moderate warm humid climate, characterized by a small
annual range of relatively high mean monthly temperatures and a distinguishable dry period
(Figure 2.2).
2.3.1 Temperature
The mean annual temperature is 25.5/C with a low seasonal variation, since the
oscillation of the mean monthly temperatures through the year is less than 5/C (INEGI, 1994).
Besides sun radiation, temperature is influenced by changes in wind and humidity. The coolest
period occurs from December to February, when cold winds from the north (nortes) affect the
island and the humidity is moderately low. The hot period occurs from May to September, when
10
high temperatures and humidity are induced by an increase in solar radiation and the action of
the highly humid and weak warm winds from the Atlantic Ocean and the Caribbean Sea.
FIGU RE 2.2 Mean monthly temperature (dots) and rainfall (bars) in Cozumel Island (data from 44 years of
observation -INEGI, 1994).
2.3.2 Rainfall
The mean total annual rainfall in Cozumel Island is 1,505 mm (INEGI, 1994). The
seasonality in Cozumel Island is due to the irregularity in the rainfall distribution through the
year. During the hot months, the wind circulation pattern transports water vapour from the
Atlantic Ocean and the Caribbean Sea to the continent, causing rainfall. This rainy season
extends from late May to October, with two maximum periods of rainfall, one in June and the
other in September-October, with a brief relative dry period between them in July and August.
Rainfall during this period is the most abundant in the year, and is associated with the formation
of tropical storms and hurricanes. In the cool months, the dry and cold fronts from North
America move southwards, becoming slightly humid when they cross the Gulf of México, before
11
reaching the Yucatán Peninsula and Cozumel Island; b ecause of this, from November to January,
there is a period of additional rainfall, albeit lower in intensity.
2.4 HURRICANES
The hurricanes that strike the Caribbean coast of the Yucatán Peninsula originate in the
Caribbean Sea and in the African Atlantic. The hurricane season in the Yucatán Peninsula runs
from May to November, i.e., basically during the rainy season; September being the month with
the highest frequency and intensity of hurricanes.
The northern half of the Caribbean coast of the Yucatán Peninsula, where Cozumel Island
is located, is the area most frequently struck by hurricanes in México. The probability of a
hurricane in the Caribbean coast of the Yucatán Peninsula is one every 3.5 years, but the
occurrence of a super-hurricane (winds >280 km/h) is one every 100 years (Morales, 1993).
Twenty hurricanes struck Cozumel Island from 1871 to 1995 (SPP, 1979; Jáuregui et al., 1980;
Antochiw and Dachary, 1991; Merino-Ibarra and Otero-Dávalos, 1991; Morales, 1993; Landsea,
1996), giving an overall frequency of one hurricane every 6.2 years, and a frequency of one
hurricane of category 5 (winds>248km/h, atmospheric pressure<920 millibars) every 124 years
(Table 2.1).
Hurricanes are probably one of the most important ecological factors affecting the
structure and composition of forests in the Caribbean region. The differential mortality, recovery
from damage, growth and recruitment can induce radical changes in plant population abundances
and in the physical structure of the forests, as reported in other forests of the Caribbean region
affected by hurricanes (Crow, 1980; Lugo et al., 1983; Brokaw and Walker, 1991; Guzmán-
Grajales and Walker, 1991; Walker, 1991; Yih et al., 1991; You and Petty, 1991).
The huge quantities of leaves, branches and dead logs left after a hurricane, and dried
during the following dry season, make forests susceptible to wildfires, which can have a very
important influence in forest structure and may have more important effects than direct hurricane
damage as reported in the mainland of the Yucatán Peninsula (Whigham, et al., 1991; Morales,
1993). Although important in the mainland, wildfires have not been important at Cozumel
Island, according to the available information.
12
TAB LE 2.1 Hurricanes that have passed over or near Cozumel Island from 1871 to 1995, showing their
characteristics when passing over or near the island (see text for sources).
Year M onth
Hurricane
Name Category
Atmosph.
Pressure Intensity Location
1879 September <10km NE off Cozumel Island
1880 August .20km NE off Cozumel Island
1886 June 1 148 km /h .10km E off Cozumel Island
1886 June 2 157 km /h .50km SW off Cozumel Island
1887 July 2 157 km/h Cozumel Island
1887 September 2 157 km /h .40km NE off Cozumel Island
1893 September 2 157 km /h .50km NE off Cozumel Island
1903 August 2 176 km/h Cozumel Island
1912 October 1 148 km/h .20km N off Cozumel Island
1913 June 2 157 km/h <30km SW off Cozumel Island
1922 October 1 130 km/h <10km NE off Cozumel Island
1933 September 2 167 km /h .10km S off Cozumel Island
1938 August 2 157 km/h <30km E off Cozumel Island
1938 August 2 157 km/h .20km SE off Cozumel Island
1942 August 2 167 km/h .20km N off Cozumel Island
1944 August 1 148 km/h Cozumel Island
1951 August Charlie 4 964 mb 213 km/h Cozumel Island
1967 September Beulah 2 970 mb 176 km/h Cozumel Island
1988 September Gilbert 5 892 mb 268 km/h Cozumel Island
1995 October Roxanne 3 958 mb 185 km/h Cozumel Island
2.5 VEGETATION
The vegetation of Cozumel Island (Figure 2.1) is similar to that of the eastern coast of the
Yucatán Peninsula, made up of warm climate plant associations (Téllez-Valdés, et al., 1989).
The main vegetation types are tropical semi-deciduous forest, low tropical deciduous forest, and
13
mangrove forest. These are the vegetation types greater in area and their distribution is
determined mainly by the soil.
Other types of vegetation are the tasistal (a palm association, see below), the cattail
swamp and the halophilus dune vegetation, which are lesser in area and can occur between the
main vegetation types.
Secondary vegetation replaces the primary ones when these are partially or totally
transformed by hurricanes, fire or human intervention. The secondary associations tend to revert
to primary after some time, depending on the severity and the duration of the alteration.
2.5.1 Tropical Semi-deciduous Forest
This vegetation type consists of two strata from 8-20 m high, and a limited shrub-
herbaceous understorey. The shrub stratum is physiognomically and floristically better defined
towards the centre of the island. There are a few epiphytes and climbers. Some of the dominant
tree species of this community are Manilkara zapota (sak-ya, zapote de monte, chicozapote or
chicle), Bursera simaruba (chacah), Calliandra belizensis (barbas de viejo), Cedrela odorata
(cedro), Metopium brownei (chechem), Vitex gaumeri (ya'axnik), Ceiba pentadra (ceiba),
Lysiloma latisiliqua (tzalam), and Mastichodendron foetidissimum (caracolillo). This forest
grows on a shallow soil with little organic matter (Cabrera-Cano et al., 1982; Téllez-Valdés and
Cabrera-Cano, 1987; Téllez-Valdés, et al., 1989). The tropical semi-deciduous forest represents
the most extensive vegetation type of the island, occupying approximately 286 km . In the areas
2
close to the cenotes, more trees retain their leaves along the year, giving the impression of a
tropical semi-perennial forest. Furthermore, in this areas Brosimum alicastrum (ox or ramón)
becomes a common tree species.
2.5.2 Low Tropical Deciduous Forest
This community is usually composed of a main arboreal stratum 8-12 m high, and a shrub
stratum, without a herbaceous understorey. Epiphytes and climbers are scarce. In areas
susceptible to seasonal flooding its composition is complemented by other elements, and there
are more epiphytes and climbers. Some of the most important elements of this community are
Enriquebeltrania crenatifolia (chiim took), Pithecellobium mangense (chakchucum), P. dulce
(guamuchil), and Diospyros nicaraguensis (uchiche') (Cabrera-Cano et al., 1982; Téllez-Valdés
14
and Cabrera-Cano, 1987; Téllez-Valdés, et al., 1989). This forest surrounds the tropical semi-
deciduous forest and is located mainly in the perimeter of the island, just behind the swamp
vegetation, the halophilus dune vegetation, and the mangroves (Figure 2.1). This forest type has
an area of approximately 63 km .
2
2.5.3 Mangrove Forest
This is one of the most characteristic coastal vegetation communities of the tropics. In
Cozumel Island it consists mainly of arboreal elements 5-10 m high, dominated by Rhizophora
mangle (tabche or mangle rojo), Laguncularia racemosa (mangle blanco), Conocarpus erectus
(mangle botoncillo) and Avicennia germinans (mangle negro). It may have some epiphytes and
climbers, such as the orchids Brassavola nodosa (awoche') and Schomburgkia tibicinis, the
cactus Selenicereus testudo (pitaya), the bromeliad Aechmea bracteata (chal-ka'nalsihil or xchu),
and some Apocynaceae species. This community is characterized by a poor tree species
diversity, high humidity and temperature, and soil rich in organic matter, periodically to
permanently flooded in brackish to salt waters. Occasionally Manilkara zapota and Annona
glabra (árbol de corcho) tolerate these conditions (Téllez-Valdés and Cabrera-Cano, 1987;
Téllez-Valdés, et al., 1989). The most important mangrove communities in Cozumel Island are
found in the northern and southern salt lagoons (Figure 2.1). This vegetation type occupies
approximately 31 km of the island.
2
2.5.4 Other Vegetation Types
The tasistal is a low-diversity palm association dominated by Acoelorrhaphe wrightii
(tasiste). It grows on periodically or permanently flooded soil, consequently it can be found next
to the mangrove forest or in aguadas inside the forests.
The cattail swamps are associated with periodically or permanently flooded and muddy
soil. They are usually single species dominated communities of Typha dominguensis (tule) or
Cladium jamaicense. This vegetation is found next to the mangrove forest and in the interior
lagoons around the island that are formed during the rainy season.
The halophilous dune vegetation is composed mainly of erect and postrate plants with
shrubby and herbaceous life-forms which are exposed to strong winds, high salinity and sun
15
radiation. It is found on sandy, gravelly or rocky soils with little organic matter. Its floristic
composition varies notably, depending on the locality on the island (Téllez-Valdés, et al., 1989).
2.5.5 Secondary Vegetation
These are communities which develop when primary vegetation (mainly tropical semi-
deciduous and low tropical deciduous forests) is totally or partially transformed. They are
composed of some arboreal strata between 5-12 m high, several shrub understoreys and a
herbaceous one, with many epiphytes and climbers. These associations cover basically the areas
influenced by human activities, such as abandoned agricultural lands, roadsides, paths and
settlements. They are also found in areas of natural disturbance caused by hurricanes and fire.
The common species of this vegetation type are pioneers that have high dispersal efficiency,
rapid growth and sometimes resistance to fire. The most common species are Cecropia
obtusifolia (k'oochle), Byrsonima bucidaefolia, Trichilia havanensis (sakch'obenche') and
Leucaena leucocephala (huatsim) (Téllez-Valdés and Cabrera-Cano, 1987; Téllez-Valdés, et al.,
1989). These communities occupy approximately 44 km of the island.
2
2.6 ENDEMIC VERTEBRATE FAUNA
Cozumel Island supports some endemic vertebrate species owing to its isolation from the
mainland. It is difficult to present a comprehensive list of endemic vertebrate species
(particularly in the case of fishes) because of the lack of studies and also, because of recent
changes in the taxonomic position of some of them. Only birds and mammals are considered
here, as an overview of the endemic taxa of Cozumel Island. Some endemic reptile subspecies
occur (Taylor and Cooley, 1995), but no endemic amphibians have been reported for the island
(Lee, 1996).
Birds are the best-known group of vertebrates in Cozumel Island (Table 2.2). Four
endemic species and 15 endemic subspecies are now recognized (Howell and Webb, 1995).
Three endemic species of mammals were formerly recognized, but the Cozumel coati
(formerly Nasua nelsoni) was recently placed as a subspecies of Nasua narica (Decker, 1991).
16
Thus, two endemic mammal species are now recognized and four endemic subspecies (Hall,
1981; Wilson and Reeder, 1993), as shown in Table 2.3.
TAB LE 2.2 Endemic birds of Cozumel Island.
Endemic taxon Family Common name
Species
Chlorostilbon forficatus Trochilidae Cozumel emerald
Troglodytes beani Troglodytidae Cozumel wren
Toxostoma guttatum Mimidae Cozumel thrasher
Vireo bairdi Vireonidae Cozumel vireo
Subspecies
Buteo magnirostris gracilis Accipitridae Roadside hawk
Crax rubra griscomi Cracidae Cozumel curassow
Centurus pygmaeus pygmaeus Picidae Yucatán woodpecker
Centurus aurifrons leei Picidae Golden-fronted woodpecker
Attila spadiceus cozumelae Tyrannidae Bright-rumped attila
Myiarchus yucatanensis lanyoni Tyrannidae Yucatán flycatcher
Myiarchus tyrannulus cozumelae Tyrannidae Brown-crested flycatcher
Polioptila caerulea cozumelae Sylviidae Blue-grey gnatcatcher
Dumetella glabrirostris cozumelana M imidae Black catbird
Cyclarhis gujanensis insularis Vireonidae Rufous-browed peppershrike
Dendroica petechia rufivertex Emberizidae: Parulinae Golden (yellow) warbler
Spindalis zena benedicti Emberizidae: Thraupinae Strip-headed tanager
Piranga roseogularis cozumelae Emberizidae: Thraupinae Rose-throated tanager
Cardinalis cardinalis saturata Emberizidae: Cardinalinae Northern cardinal
Tiaris olivacea interm edia Emberizidae: Emberizinae Yellow-faced grassquit
17
TAB LE 2.3 Endemic mammals of Cozumel Island
Endemic taxon Family Common name
Species
Reinthrodontomys spectabilis Muridae: Sigmodontinae Cozumel Island harvest mouse
Procyon pygmaeus Procyonidae Cozumel Island raccoon
Subspecies
Oryzomys couesi cozumelae Muridae: Sigmodontinae Marsh rice rat
Peromyscus leucopus cozumelae M uridae: Sigmodontinae White-footed mouse
Nasua narica nelsoni Procyonidae Cozumel Island coati
Pecari tajacu nanus Tayassuidae Collared peccary
2.7 HUMAN BACKGROUND ON THE ISLAND
Cozumel Island was first inhabited by the Maya around BC 300. There is evidence
indicating the island's continuous human habitation until the Spanish arrival in 1518. The period
between BC 300 and AD 700 was characterized, not only by a small human population size, but
also by a relatively poorly developed culture compared with the Maya from the mainland.
During the Late Classic Maya period (AD 600-900) there was an increase in the population of the
island, owing to a change in the commercial system in the Maya region from terrestrial to
maritime. The Late Post-Classic Maya period (AD 1200-1517) was a time of splendour for the
Cozumel Maya; it was when most of the main Maya buildings on the island were built, as well
as the sacbeob (stone-lined roads), and the population reached a maximum of 8,000-10,000
inhabitants (Sabloff and Rathje, 1975; Freidel and Sabloff, 1984; Antochiw and Dachary, 1991).
After the Spanish arrival in 1518, Cozumel Island preserved for a time its commercial and
religious importance for the Maya. However, the conquest of the Yucatán Peninsula modified
the ancient commercial pattern and the religious pilgrimages to Cozumel, bringing about a
reduction in population immigration and an increase in emigration. Additionally, the
introduction of smallpox by the Spanish soldiers in 1520 caused a further reduction in the Maya
18
population of Cozumel Island (Figure 2.3). By 1549, there were about 1,000 inhabitants in the
island, by 1570 this number was reduced to 400, and by the middle 17 century the island was
th
practically uninhabited.
During the time the island was deserted, it was sporadically occupied by English pirates.
Additionally, people from the oriental coast of the Yucatán Peninsula, Belize and Honduras went
to Cozumel Island only to exploit its marine and forest resources.
FIGU RE 2.3 Human population size in Cozumel Island from 1518 to 1995. T he island was practically deserted
for about 200 years (from the mid-1600s to the mid-1800s).
In 1847, when the Guerra de Castas (castes war) between the Maya and mestizos
(mixture between Amerindians and Spanish) started in the Yucatán Peninsula, some people
looked for refuge in Cozumel Island and Isla Mujeres. Hence, resettlement of the island started,
after about 200 years of almost complete abandonment. In 1849, there were already 350
inhabitants, whose economic activity was basically the agriculture, although fishing, commerce
and the exploitation of forest resources were regular activities as well. By the end of the 19th
century, the population had grown to about 1,000 inhabitants distributed in two main settlements,
San Miguel and El Cedral, and in many farms. In 1902 the Guerra de Castas ended, and in the
19
first half of the 20 century Cozumel Island became an important port. In the first decades of this
th
century, commerce of island crops was the fundamental economic activity. In the 1920s, the
exploitation and commercialisation of chicle gum (a latex harvested from Manilkara zapota
trees) in the forests of the Yucatán Peninsula and Cozumel Island gave a great economic surge
to the island until the mid-1950s. By 1960, the population was 7,562 inhabitants (Dachary and
Arnaiz-Burne, 1988; Arnaiz-Burne, 1988; Antochiw and Dachary, 1991).
From 1960s the economic activity shifted to tourism. Agriculture, fishing and the
exploitation of forest resources practically disappeared as economic activities. This brought
about the concentration of the population in San Miguel, in the west coast of the island. Tourism,
as the main economic activity, has caused a large increase in the population of the island, mainly
because of the immigration of people from the mainland of the Yucatán Peninsula, México City
and the state of Veracruz to cater for tourism services and infrastructure development. In 1995
the human population size of Cozumel Island was of about 50,000 inhabitants. The Plan
Director Urbano de Cozumel limits the population size to approximately 68,800 inhabitants by
the year 2000 (Castro-Sariñana, 1988; Antochiw and Dachary, 1991; INEGI, 1994).
20
3
METHODS
3.1 FIELD DATA COLLECTION
Field work was carried out in the tropical semi-deciduous and low tropical deciduous
forests of Cozumel Island from October 1994 to June 1995, that is, from about the end of the
rainy season of 1994 to the beginning of the rainy season of 1995.
I used line transect sampling to estimate C. r. griscomi density and population size on
Cozumel Island. A total sampling effort of 386 km along transects was invested during the
study, consisting of 367 km of tropical semi-deciduous forest and 19 km of low tropical
deciduous forest. I surveyed ten line-transects measuring from 2.12 to 5.80 km, with 20 minutes
stops every 200 metres, from about 0700 to 1600 h, on fifteen days each month. Eight of the
transects were placed on available line trails through the forests of the island, which were cleared
to delimit private properties, communal land boundaries (ejido El Cedral) and the extent of the
land under the administration of the Commission for Drinkable Water and Sewerage (Comisión
de Agua Potable y Alcantarillado - CAPA). Additionally, one of the transects was a path cleared
by people from El Cedral for access to the forest interior. The other transect, located in the north
of the island, near San Gervasio, was cleared a few years before, during archaeological studies
(Figure 3.1).
For each curassow detected during line-transect sampling, I recorded sex, age (when
possible), social status (solitary or paired), and the perpendicular distance to the line transect, in
order to be able to estimate density. Additionally, the exact position of the bird along the transect
(transects were tagged every 200 m) was registered to relate its location with habitat
characteristics.
IZ
"(SUO!1EAJOsqo
Play
pUR sa:)Jnos
(lUaAaS
WOJj pal!dW03
U0!1RUlJOJU¡)
'U"'\04S aSIR aje
IX;)l
041 U! pauopuaw SpCOJpUR
Sal!S
'sa!l!UnWW03
UO!lCla~;;¡A
JO Sl!W!1
041 pue
'S<1JOUiJ:J
lUC1Jodw!
lSOUl
:~nn
JO uoqc:J01 aljl 'paÁa..uns spasuc.Q aLJl JO uopu:JO[ alJl fiU!MOqS
'puclsI
fOwnzoJ JO dulr'J:
dll "snJa<l
(<1
B1qwol0:) uungu, (6
ugal:) Bungu,
(S
VdV:)
(L
JU~UeIBd(9
elSIAuudng (~
IBJpa:) 13
(17
OISBAJao uUS (S
~udUIX
(e:
sUl0]AI Ulund (1
S3llS
allS
a¡oua:)
llllll[ adÁI uOllul"ga/\
pasuuJl
peOJ)JIO
PUO.lpaAud
•
•
V3S NV388J1IV:)
.OOoLS
N
22
To analyse vegetation characteristics, I surveyed 141 plots (200 x 40 metres), along eight
of the transects. In each plot, I recorded the abundance of trees with dbh $ 16 cm of the
following species: Brosimum alicastrum (ramón), Manilkara zapota (chicozapote) and
Mastichodendron foetidissimum (caracolillo). These tree species were selected because of their
importance as food resources for C. rubra (Sermeño-Martínez, 1986; Jorgenson, 1993; personal
observations). The trees registered were classified in the following size categories according to
their dbh measurements: 16 cm # Size 1 < 29 cm # Size 2 < 48 cm # Size 3 < 80 cm # Size 4.
3.2 ADVERSE FACTORS DATA COLLECTION
The impact of natural and anthropogenic adverse factors was assessed by interviewing
informally local people, by doing direct observations and gathering information from a
bibliographic review. The main informants were fourteen residents of Cozumel Island who were
born in the island (50-80 years old) or had lived there for at least 25 years, but further information
was obtained from seven other people of more recent residence.
Informants provided anecdotal data about trends in abundance and the historical
distribution of C. r. griscomi on the island in the last decades. The information gathered from
informants about natural adverse factors took into account mainly the occurrence and intensity
of the strongest hurricanes that have struck the island since the 1930s. The information about
the anthropogenic adverse factors included hunting, exploitation of forest resources, land-use
trends, and the introduction of exotic fauna. The information about hunting was not easy to
obtain from informants, since hunting is an illegal activity on the island nowadays.
I also collected data by observation of hunting activities, the occurrence of wildfires and
human-induced fires, land-use patterns, and the exploitation of forest resources.
The bibliographic review included a comprehensive compilation of the data available
about the hurricanes that have struck Cozumel Island since 1871, and the gathering of statistical
and historical information about human population growth and exploitation of forest resources.
23
3.3 DATA ANALYSIS
3.3.1 Density and Population Size Estimation
Density estimation was based on the distance sampling theory (Buckland et al., 1993).
According to this theory, only a proportion of the objects (n) in a surveyed area (of length L and
a
width 2w) is detected. If that proportion (P) can be estimated from the distance data, the
estimate of density could be written as:
a
D = n / L2wP
a
The estimation of P from the perpendicular distances, results in the parameter µ , which is a
function of the measured distances (x) and represents the effective half-width of the strip
extending either side of the transect:
0
µ = I g(x)dx
w
Thus, the general estimator of density for line transect sampling is:
D = n / L2:
Field data gathered on the C. r. griscomi detection distances were analysed using the
program DISTANCE V2.2 (Laake et al., 1994). The data were treated ungrouped and 6.25% of the
farthest observations were truncated to facilitate modelling of the data. The detection function
selected by the programme based on Akaike's Information Criterion was a model with a uniform
key function and cosine series expansion. The effective strip width (or more strictly, half-width)
obtained to estimate the density of the Cozumel curassow was 23.8 metres.
Density was also estimated based on the mean perpendicular detection distances (0)
(Overton, 1971):
D = n / L20
This estimation was done only for comparison purposes with other density estimates.
Although this method is commonly used for the density estimation of cracid populations (Silva
and Strahl, 1991), it is not as robust as the density estimation based on the distance sampling
theory proposed by Buckland et al. (1993).
The density estimated with the uniform detection function was used to calculate the
Cozumel curassow population size in the island, considering only the areas of the tropical semi-
deciduous forest and low tropical deciduous forest, since these vegetation types represent the
most suitable habitat for this cracid. Areas of each vegetation type were calculated from a
24
vegetation map obtained from Museo de la Isla de Cozumel, which was modified according to
field observations.
3.3.2 Assessment of Cenotes Influence on the Cozumel Curassow Population
To evaluate the influence of cenotes and aguadas on the abundance and distribution of
C. r. griscomi, I used line transect data to estimate curassow densities for the areas within
different radial distances from those water sources, and compared those desnsities with the
densities for the areas away from each of those radial distances (Figure 3.2). The density
estimates were plotted and curve fitted by polynomial regression analysis to explore the effect
of those landscape features on C. r. griscomi abundance and distribution. The limit of the
cenotes and aguadas influence was deduced based on the visual analysis of the plots.
FIGU RE 3.2 Schematic representation of the criteria used to evaluate cenotes and aguadas influence on Crax
rubra griscomi population along line transects. Two situations are shown: 1) Transect passes on or by a cenote or
aguada, and 2) Transect does not pass by a cenote or aguada. A indicates the transect portion within r radial
distance from those water sources, and B the transect portion away from the r radial distance. A' and B' are the
respective transect p ortions for a different radial distance, r'. Radial distance was increased at 250 metres intervals
in both situations.
25
Additionally, differences between the number of curassows (as a whole and by sex)
detected near (#250 metres) and far (>250 metres) from cenotes were analysed with G-tests
adjusted with Williams's correction (Sokal and Rohlf, 1995), with the expected frequencies
adjusted to sampling effort.
3.3.3 Assessment of Human Influence on the Cozumel Curassow Population
Human influence on the Cozumel curassow population was assessed based on linear
distances from human settlements and human accesses to the forest interior (accesses to forest
trails), because the only way to move through the forests is by trails, since vegetation in the
understorey of the forests is very dense. Density estimates at different linear distances from the
sources of human influence were used to evaluate its effect on the abundance and distribution
of this cracid. The densities estimated were plotted and curve fitted by polynomial regression
analysis to find differences in the levels of human influence on the Cozumel curassow at different
distances from the sources of influence. Additionally, the exact detection distance from the
sources of human influence, of all C. r. griscomi sightings (as a whole and by sex) were also
analysed to find further evidence of the levels of human influence.
3.3.4 Assessment of Vegetation Influence on the Cozumel Curassow Population
To assess the influence of the abundance of Brosimum alicastrum, Manilkara zapota and
Mastichodendron foetidissimum on the abundance and distribution of C. r. griscomi, I compared
the abundances of these tree species in plots where this cracid was detected and those where it
was not detected. Statistical comparisons were made with the Mann-Whitney U-test corrected
for ties (Sokal and Rohlf, 1995). To avoid any bias caused by the relation curassow-cenote, I
eliminated from the analysis, the plots within a radius of 250 metres from a cenote.
26
4
RESULTS
Seventeen observations of C. r. griscomi were recorded in 113 samples (transects
surveyed), within a total of 386 km of forest surveyed along the transects. The observations
included nine male and eight female detections, which represented six male and seven female
curassows individually identified (Table 4.1). All detections occurred in the tropical semi-
deciduous forest, and no Cozumel curassow was detected in the low tropical deciduous forest,
although this cracid also inhabits this vegetation type according to local informants. The lack
of detections in this forest type is likely to be related to the difference in sampling effort between
the two forest types, and also probably with the difference in their ecological importance for the
Cozumel curassow as well as the degree of human influence.
4.1 DENSITY OF THE COZUMEL CURASSOW
The estimated C. r. griscomi density on Cozumel Island, or more specifically, in the
tropical semi-deciduous and low tropical deciduous forests of the island, was 0.87 birds/km ,
2
with a relatively wide confidence interval because of the small number of detections (Table 4.2),
an inherent difficulty when working with endangered or rare taxa. This density represents a
biomass of approximately 2.7 kg/km , according to the body mass reported for C. r. rubra in the
2
mainland of Yucatán Peninsula (3.1 kg) by Jorgenson (1993, 1995), but it is reasonable to believe
that C. r. griscomi biomass is slightly lower, since this curassow is smaller than the one on the
mainland.
If only the tropical semi-deciduous forest is considered, the density of the Cozumel
curassow was 0.91 birds/km (Table 4.2), which suggests a slight preference for this vegetation
2
type, probably because of the higher fruit production and the presence of more and larger cenotes
in this forest. The higher density may also suggest a relatively lower human influence in the
27
tropical semi-deciduous forest, because of a greater inaccessibility if compared with the low
tropical deciduous forest (see below).
The estimated density of the Cozumel curassow based on the mean perpendicular
distances method (Overton, 1971) was 1.7 birds/km . This value is very likely to represent an
2
overestimate of the density, thus care has to be taken when considering densities estimated using
this method (e.g. Silva and Strahl, 1991).
TAB LE 4.1 Crax rubra griscomi detections in the forests of Cozumel Island during October 1994 - June 1995.
Sex, age class, detection time and perpendicular detection distances to line transects of birds detected on each
study period are shown, as well as the number of samples (transects surveyed) and sampling effort invested on
each.
Period
Number of
samples
Sampling
effort (km) Sex Age class
Detection
time
Perpendicular
distance (m)
October-November 18 63.51 No detections
November-December 19 60.4 No detections
December-January 12 45.9 %Adult 1240 1
January- February 16 59.71 &
&
&
?
?
?
0914
0914
1050
2.5
2.5
10.0
February-March 14 43.68 %
&
%
%
Adult
Adult (?)
Adult
Adult
1300
1300
0859
1344
16.0
16.0
23.0
16.0
March-April 6 21.31 &
%
?
Adult
1345
1628
23.8
6.0
April-May 19 63.25 %
%
%
Subadult
Adult
Adult
1327
1414
1030
1.0
16.0
1.0
May-June 9 28.35 &
&
%
&
Adult (?)
Adult (?)
Adult
Adult (?)
0851
1150
1400
1420
21.0
29.8
17.0
17.0
28
4.2 POPULATION SIZE OF THE COZUMEL CURASSOW
Considering the calculated density and the 349 km of tropical semi-deciduous and low
2
tropical deciduous forests of the island, which represent the main potential suitable habitat for
the Cozumel curassow, a population of 304 curassows inhabiting Cozumel Island was estimated,
where the tropical semi-deciduous forest would be supporting most of the population (Table 4.2).
TAB LE 4.2 Density, population size and encounter rate estimates of Crax rubra griscomi in Cozumel Island
forests (349 km ) and in the tropical semi-deciduous forest only (286 km ).
22
Parameter Estimate Standard Error
% Coef.
Variation
Confidence Interval
(95%)
Cozumel Island Forests
Density (birds/km ) 0.87 0.28 32.12 0.47 - 1.61
2
Population size (birds) 304 97.64 32.12 164 - 562
Encounter rate (birds/km) 0.04 0.01 32.12 0.02 - 0.08
Tropical Semi-deciduous Forest
Density (birds/km ) 0.91 0.30 32.29 0.49 - 1.70
2
Population size (birds) 262 84.61 32.29 141 - 485
Encounter rate (birds/km) 0.04 0.01 32.29 0.02 - 0.08
If the sex ratio is estimated based on birds detected (Table 4.1), the proportion of
males:females is 1.00:0.89, but if the estimate is based only on individually identified birds, since
six males and seven females were observed, the sex ratio is 1.00:1.17. Thus, sex ratio in the
Cozumel curassow population is likely to be close to 1:1.
Although males can be age-identified (adult, subadult), it is not possible to identify
females' age from sightings in the field. In the case of males, only one of the six individually
identified males was subadult (based on the size of the knob on the bill). Considering only the
age proportion of males, it could be roughly assumed that the C. r. griscomi population averages
29
about 253 adults and 51 subadults (#1 year old). Thus, the population of the Cozumel curassow
probably consisted of approximately 126 adult males, 126 adult females, 25 subadult males and
25 subadult females.
Table 4.3 Life and fertility table of C. r. griscomi, given the available information. The life-span (seven years),
the survivorship values and the age of sexual maturity are supposed (see text for details). The fertility value is
assumed to be the same for all age classes after sexual maturity. The terminology used was adapted from Pielou
(1974).
Age class in
years (x)
Age-specific
x
survival rate (l)
Age-specific
x
fertility (m)
Reproductive rate of
xx xx
each age class (lm)(xl m )
0<x#1 0.5 0 0 0
1<x#2 0.5 0 0 0
2<x#3 0.5 0 0 0
3<x#4 0.7 0.2 0.14 0.56
4<x#5 0.7 0.2 0.14 0.70
5<x#6 0.7 0.2 0.14 0.84
6<x#7 0.7 0.2 0.14 0.98
0xx
Net reproductive rate = R = 3lm = 0.56 females/female/generation
xx 0
Mean length of a generation = G = 3xl m ÷ R = 5.5 years
0
Instantaneous rate of population increase = r = ln R ÷ G = -0.11 curassows/curassow/year
xx
when 3lme=1
-rx
Finite rate of population increase = 8 = e = 0.90 curassows/curassows/year
r
The sex ratio of C. r. griscomi population suggests that males were sufficiently numerous
to ensure that no female went unmated; thus the instantaneous rate of population increase (r)
could be calculated from the female part of the population. Given the available data, fertility rate
(m) was estimated in approximately 0.2 subadult females/adult female/year. If it is assumed that
50% of female curassows will survive to age of first breeding (Silva and Strahl, 1991), then the
survivorship is 0.7 at all ages (a liberal estimate). If it is also assumed that fertility is the same
30
for adult females, independently of the specific age, that sexual maturity is reached at the age of
three years old (J. Estudillo, cited in Silva and Strahl, 1991), and that the Cozumel curassow has
a life-span of seven years (a liberal estimate, but see Appendix II), then the instantaneous rate of
population increase would have a value of about r=-0.11 curassows/curassow/year. Similarly,
the finite rate of increase would be 8=0.90 curassows/curassow/year (Table 4.3).
If different life spans are considered, drastically different values of the instantaneous rate
of population increase are obtained. With a life span of only 5 years, the population would be
declining at a rate of r=-0.28 curassows/curassow/year. With a life span of 10 years, the
population would be slightly below equilibrium (r=-0.003 curassows/curassow/year), and with
a life span greater than 10 years the population would be increasing (see Appendix II).
4.3 CENOTES INFLUENCE ON THE COZUMEL CURASSOW
POPULATION
Because C. r. griscomi detections occurred mainly during the dry season, the results
obtained here clearly show the cenotes and aguadas influence on the Cozumel curassow
distribution. The cumulative density estimates obtained at different radial distances from these
water sources are reported in Appendix III. The highest densities of this endemic cracid in the
forests of the island were found near the cenotes and aguadas, and they decreased markedly
further from these water sources (Figure 4.1A). When comparing this plot with its counterpart
(Figure 4.1B), i.e., the plot of C. r. griscomi density away from a given distance from the cenotes
and aguadas, a striking difference can be seen between the curves, but this difference is lost at
about 2 km from the cenotes and aguadas. Thus, this suggests that the greatest cenotes and
aguadas influence on C. r. griscomi density could be within 2 km of radial distance from those
water sources.
31
FIGU RE 4.1 Influence of cenotes on the Cozumel curassow density. A) Circles (with standard error bars)
represent the cumulative density at different radial distances from cenotes. The continuous line defines the best
curve fitted for this cumulative density (y=7.24 -9.976x+6.075x -1.204x ; r =0.98; p<<0.001). B) Squares (with
232
standard error bars) represent the cumulative density away from a given radial distances from cenotes. The dotted
line defines the best curve fitted for this cumulative density (y=0.638-0.484x-0.004x +0.205x ; r =0.91; p<0.002).
232
The analysis of the location of the curassows detected with respect to the cenotes and
aguadas, showed a clear cluster of birds near these water sources. Almost 60% of all curassows
detected were found near cenotes and aguadas, although there was a sex bias in this clustering,
with 88% of the females and 33% of the males detected. This clustering is even more striking
when analysing the number of curassows detected near (#250 m) and far (>250 m) from cenotes
and aguadas considering the sampling effort (36 km surveyed near and 350 km far from cenotes
and aguadas) (Figure 4.2). There were significantly more curassows detected near cenotes and
adj
aguadas (G =25.039; df=1; p<<0.001) than far from them. Analysing by sex, both males
32
adj adj
(G =21.088; df=1; p<<0.001) and females (G =25.764; df=1; p<<0.001) were more often seen
near cenotes and aguadas (Figure 4.2).
FIGU RE 4.2 Crax rubra griscomi encounter rate (curassows detected per kilometre of transect surveyed) in the
tropical semi-deciduous and low tropical deciduous forests, at 0.25 km intervals of radial distance from cenotes.
4.4 HUMAN INFLUENCE ON THE COZUMEL CURASSOW
POPULATION
Human settlements (i.e., urban and rural settlements), human accesses to the forest
interior (roads and trails), and human activities within the forests (e.g., hunting, wood gathering),
were found to have a negative effect on the Cozumel curassow abundance and distribution
(Figure 4.3 and 4.4).
Three levels of human influence were identified (Figure 4.3, Appendix IV). The first
occurs between 0 to 1.5 km from human settlements and accesses, where the Cozumel curassow
33
is absent. The second level arises between distances >1.5 to 3 km, where the density of this
cracid is depleted with respect to the overall density. The last level happens after 3 km from
human settlements and accesses, where density tends to achieve the value of the overall
population density on the island.
FIGU RE 4.3 Human influence on the Cozumel curassow density. Circles represent the cumulative density at
different distances from human settlements and human accesses to the forest interior, showing standard error bars.
The continuous line represents the best curve fitted and the dotted lines define the 95% confidence interval of the
cubic polynomial regression curve (y=-0.192+0.173x +0.027x -0.004x ; r =0.92; p<0.001).
232
A further analysis of the Cozumel curassow density beyond 3 km from human settlements
and accesses apparently splits this sector in two levels of human influence (Figure 4.4, Appendix
IV). One level is the interval between >3 to 4.5 km, with a lower population density than the
interval >4.5 to 6 km. The Welch's approximate t-test, however, shows no statistical difference
s
between the Cozumel curassow density estimates within these two intervals (t' =0.59; df=87;
p>0.5).
34
FIGU RE 4.4 Cozumel curassow density at 1.5 km intervals from human settlements and accesses to the forest
interior. Bars represent density within each interval, showing the standard error.
The exact detection distances from human settlements and human accesses to forest trails
of all C. r. griscomi sightings (Figure 4.5), suggest that the limit of the highest human influence
is located at about 1.9 km, since the nearest detection occurred at that distance from human
settlements and accesses to the forest interior. It should be noted that 91% of the detections
between distances 3.2 and 4.4 km from human influence were associated with cenotes and
aguadas. Thus, it can be assumed that the occurrence of the Cozumel curassow at this distance
interval is influenced by the presence of these water sources, not only because of a decrease in
human influence. The effect of human influence can be corroborated by the fact that no
curassows were detected in an aguada located 2.7 km from human settlements and accesses. If
we only consider the sightings not associated with cenotes and aguadas, apparently the actual
decrease in human influence arises farther than 4.4 km from the sources of human influence
(Figure 4.5).
Females could be considered more sensitive to human influence, since the nearest female
was detected at 3.2 km from human settlements and accesses to the forest interior, while the male
35
closest to such sources of human influence was detected at 1.9 km (Figure 4.5). Farther than 3.2
km from human settlements and accesses, there does not seem to be any difference between sexes
regarding the distance from human influence.
FIGU RE 4.5 Exact distance from human settlements and human accesses to the forest interior of all Crax rubra
griscomi detections by sex, in the transects surveyed in the forests of Cozumel Island. The areas influenced by the
cenotes are indicated with the arrows.
4.5 VEGETATION INFLUENCE ON THE COZUMEL CURASSOW
DISTRIBUTION
The tree species that showed a significant influence on the distribution of the Cozumel
[4]12
curassow, were Manilkara zapota (t =-3.622; n =124; n =10; one-tailed p<<0.001) and
[4]12
Mastichodendron foetidissimum (t =-2.258; n =124; n =10; one-tailed p<0.012), since C. r.
griscomi detections tended to be in the plots that held the higher abundances of these tree species.
36
It was basically because these tree species represent important food resources for the Cozumel
curassow. This influence is even more striking when considering only the plots where the largest
[4]12
trees (dbh$29 cm) of these species are located (t =-4.07; n =124; n =10; one-tailed p<<0.001
[4]12
for M. zapota, and t =-2.875; n =124; n =10; one-tailed p<0.002 for M. foetidissimum).
Coincidentally, these areas are located far (>3.5 km) from human settlements and accesses to the
[4]12
forest interior (t =-4.614; n =78; n =63; one tailed p<<0.001).
On the other hand, Brosimum alicastrum did not show any relationship with the
[4]12
distribution of C. r. griscomi (t =-1.085; n =124; n =10; one-tailed p=0.14), regardless of its
importance as a food resource (Sermeño-Martínez, 1986; Jorgenson, 1993). The reason was
probably due to the fact that all detections of this curassow were between January and early June,
i.e., out of its fruiting period in Cozumel Island. The individuals of this tree species were never
seen fruiting, and I was also informed by the local people that the fruiting period of B. alicastrum
is basically during the rainy season. Additionally, because B. alicastrum is closely associated
[4]12
with cenotes (t =-4.362; n =130; n =11; one-tailed p<<0.001), the possible link between the
Cozumel curassow and B. alicastrum could have been obscured, since the plots close to cenotes
were eliminated from the analysis of the influence of the tree species on the Cozumel curassow
distribution, to avoid bias caused by the curassow-cenote relation.
4.6 NATURAL FACTORS AFFECTING THE COZUMEL CURASSOW
POPULATION
4.6.1 Hurricanes
A comprehensive review of the hurricanes that have passed over or near Cozumel Island
is presented in Chapter 2 (Table 2.1). From 1886 to 1950, the fourteen hurricanes that struck
Cozumel Island were of category 1 or 2 (but see Landsea, 1996), with maximum sustained winds
between 130-176 km/h (there are no data on hurricane intensity before 1886). From 1951 to
1995, although the frequency of hurricanes declined to less than one every decade, three of the
strongest hurricanes that have occurred in the area since 1886, passed over Cozumel Island
(Figure 4.6). In 1988, Hurricane Gilbert, the strongest cyclone ever measured in the Western
Hemisphere, struck Cozumel Island and was followed by an unusual dry season in 1989 (Lynch,
37
1991; Whigham et al., 1991; Morales, 1993). In October 1995, just after the field work of this
study was concluded, Hurricane Roxanne struck Cozumel Island. Although not as strong as
Hurricane Gilbert, it was the third strongest hurricane that hit the island since 1886.
The possible effect of hurricanes on the Cozumel curassow population are discussed later
(section 5.4).
FIGU RE 4.6 Frequency of hurricanes and intensity of the strongest hurricane in a decade that have passed over
or near Cozumel Island from 1871 to 1995. Hurricane intensity estimates are less reliable before 1950 than after
that year (Landsea, 1996).
4.6.2 Wildfires
Pérez-Villegas (1980) reported no wildfires in the eastern coast of the Yucatán Peninsula,
during the period from 1960 to 1978, except for 1975, when several fires were recorded, although
it is not specified if fires also occured in Cozumel Island. There are apparently no subsequent
reported records of fires in the area, until the dry season of 1989, after Hurricane Gilbert, when
one of the most severe wildfires occurred in the Yucatán Peninsula, affecting hundreds of square
kilometres of forest on the mainland (Lynch, 1991). There is not Information for Cozumel Island
about wildfires after Huricane Gilbert, probably they never occured or were not important.
38
On May 1995, at the end of the dry season, an important wildfire affected the north-east
coast of the Yucatán Peninsula, near Cancún, which could not be controlled until the June rains
started. At the same time, two small wildfires occurred in the north of Cozumel Island.
The evidence shows that wildfires in Cozumel Island are not as important as in the
mainland of the Yucatán Peninsula, but they could represent potential catastrophes.
4.7 ANTHROPOGENIC ADVERSE FACTORS
4.7.1 Hunting
According to the information obtained from local informants and direct observations, the
Cozumel curassow is hunted mainly during the dry season (December to May), which coincides
with its reproductive period (February to June). Hunting is carried out during this period in the
cenotes and aguadas, because these are the only places where water is available in the dry season
and, therefore, the movements of this cracid are more predictable. The Cozumel curassow is
hunted with shotguns of different gauge (12, 16, 20, 410), or with .22 calibre rifles.
According to the informants, before the 1950s the Cozumel curassow was commonly
hunted by the local people and some commercial hunters, who used to hunt at the weekends
every week or fortnight during the dry season, mainly around San Gervasio, in the area between
Xlapac and Punta Molas, El Cedral and Buenavista (Figure 3.1). Besides the cenotes and
aguadas as hunting areas, informants also reported that the Cozumel curassow was sometimes
found in big flocks, in areas dominated by Brosimum alicastrum (ramonales), where it was
hunted as well. Hunters were able to kill between 3 and 6 cracids in every incursion. One of the
informants estimated that the number of curassows killed annually by commercial hunters alone
at that time was probably about 300-400 birds.
Informants also reported that just before the strike of Hurricane Gilbert in 1988, between
15 and 20 curassows were killed by local hunters every year in the area of El Cedral in Cozumel
Island. By that time hunting was already forbidden in the island. Since then, because of the
impenetrable tangle produced in the forest by the cyclone and the reduction in the Cozumel
curassow population, the hunting harvest was reduced to 5 birds each year in El Cedral. I was
informed that, in 1994, four curassows were hunted there, and in 1995 at least one curassow was
39
killed in the same area (personal observation). The area between Xlapac and Punta Molas is still
visited by about 5 hunters from San Miguel. In Buenavista hunting has practically disappeared
because it is now a tourist area, as well as in San Gervasio, where there is an archaeological site.
In the ranches surrounding San Gervasio, however, hunting is sometimes carried out by people
from the ranches.
4.7.2 Habitat Disturbance
Before the human resettlement of Cozumel Island in 1847, its forests had already been
subjected to logging to exploit Cedrela odorata (cedro), Manilkara zapota, (chicozapote),
Guaiacum sanctum (guayacán), Diospyros sp. (ébanos) and some other tree species. It is likely
that this exploitation had increased during the resettlement of the island, but the felling of the
island forests began due to agricultural activities after the human resettlement. Agriculture was
one of the most important economic activities until the 1960s, when tourism became the
predominant economic activity. During that time, at least twenty important agricultural ranches
were established on the island, and the forest conversion rate into agricultural areas was probably
the highest for the island after the human resettlement. Additionally, about four cattle ranches
were established. Although cattle ranching was not as important as agriculture, it probably
represented a further pressure upon the island's forests (Arnaiz-Burne, 1988; Dachary and Arnaiz-
Burne, 1988; Antochiw and Dachary, 1991).
From the 1920s to the 1950s, when the exploitation of chicle gum (a latex from M. zapota
trees) was a prominent economic activity on the island (Antochiw and Dachary, 1991), the felling
of M. zapota trees was stopped. However, the creation of paths into the forest to harvest the
latex, increased access to areas otherwise unaccessible in the forest interior. Thus, it is likely that
hunting pressure on the Cozumel curassow population had also increased in that period. In the
1950s, the chicle gum market collapsed (Antochiw and Dachary, 1991), consequently the harvest
of the latex was reduced and the paths to the forest interior gradually disappeared, but the
exploitation of M. zapota trees apparently increased again for local consumption.
In the late 1960s, the economic activities of Cozumel Island shift to tourism, and the
economic growth due to this activity, furthered the human population growth in the island
(Castro-Sarimaña, 1988; Antochiw and Dachary, 1991) (Figure 2.3). Since then, the expansion
of San Miguel city, and the creation of rural settlements in the island because of population
40
growth, have been an important form of human pressure upon the island forests. Furthermore,
the establishment of new resorts represents an additional pressure upon forests and mangroves.
In the 1970s the island's paved roads were built, bringing about a substantial increase in
access to formerly relatively isolated forest areas and creating an important obstacle to the
dispersion of C. r. griscomi. The Carretera Transversal (transversal road) has practically
divided the habitat of this cracid into two fragments, to the north and south of the road. Dispersal
of the Cozumel curassow between these fragments occurs (according to informants), but is
probably low, since most of this road is surrounded by ranches and secondary vegetation, except
for a section of two kilometres, towards the east coast of the island (Figures 2.1 and 3.1). It is
also the busiest road on the island. In the same way, the Carretera Costera (coastal road) has
surely prevented free dispersal of C. r. griscomi from the forest interior to the mangroves and
coast. Thus, the construction of the island roads evidently reduced the habitat availability for the
Cozumel curassow.
Nowadays, more than 75% (380 km ) of Cozumel Island is still forested (tropical semi-
2
deciduous, low tropical deciduous and mangrove forests - Figure 2.1). Agriculture and cattle
ranching are not at all important economic activities on the island, since only about 2% of the
population in 1990 was engaged in them (Castro-Sariñana, 1988; INEGI, 1994). These activities
are basically developed in areas of secondary vegetation, representing a low degree of effect on
the present C. r. griscomi habitat, although the forests can be affected to some extent during the
slash-and-burn practices carried out in the secondary vegetation (local informant, personal
communication). Likewise, felling of trees is not allowed in the island, but illegal exploitation
of hardwood is still carried out by residents, mainly around rural settlements and ranches and
along roads, for the construction of houses, huts and fences (activities related to human
population growth) and for making charcoal. Today's main threat to the island's forests is human
population growth, as a result of the economic growth due to tourism.
4.7.3 Exotic Fauna
The exotic fauna introduced into Cozumel Island forests that may have an adverse effect
on C. r. griscomi population comprise feral dogs and cattle, and the recently introduced boa (Boa
constrictor).
41
A pack of feral dogs of unknown size is known to occur in the municipal refuse tip, which
is located in the east of the island (Figure 3.1). These feral dogs are very likely to move to the
forest interior, becoming potential predators of native wildlife, and thus of adults and chicks of
the Cozumel curassow. Additionally, dogs and cats from rural settlements and ranches, can also
venture to the forest interior, but they are normally restricted to the secondary vegetation that
surrounds such settlements. These dogs and cats are potential predators after hurricanes or fires,
when the Cozumel curassow is impelled to move more widely, sometimes out of the forest
(Paynter, 1955; local informants, personal communication).
According to some local informants, feral cattle occur in the area between Xlapac and
Punta Molas in the north-east of the island, because of the release of 20-30 head of cattle, about
50 years ago, after a hurricane (probably in 1951). The cattle may affect the habitat, modifying
the structure and composition of the forest due to stamping and overgrazing of some particular
plant species, producing a nearly total absence of successful seedling regeneration (e.g. Coblentz,
1990; Rodríguez-Estrella et al., 1996). This could have an adverse effect on the Cozumel
curassow distribution and habitat availability, through changes in habitat characteristics.
Although Boa constrictor is well known from the Yucatán Peninsula (Lee, 1996), there
were no records of this snake for Cozumel Island despite explorations by naturalists and
herpetologists since the 1800s (Smith and Taylor, 1945). The first record of the species in
Cozumel, was made until 1991 (Lee, 1996). Several informants affirmed that the boa was
introduced into Cozumel Island in 1969, during the production of the Mexican film "El jardín
de la tía Isabel". About 10 to 30 boas were released around Palancar beach and on a beach
between Xlapac and Punta Molas (Figure 3.1). Almost thirty years after the introduction of this
snake, the boa is now widespread in all the island. Its population status has never been assessed,
but during the field work, seven boas were found incidentally along the transects (plus two boas
seen during prospective surveys and one more at San Miguel), representing an encounter rate of
1.81 boas for every 100 km of forest surveyed. All the boas but one were about 1.5 metres long,
the exception being about 2.5 metres.
42
5
DISCUSSION
5.1 COZUMEL CURASSOW DENSITY AND OTHER CURASSOW
DENSITIES
C. r. griscomi ecological density (density estimate based on the suitable habitat available)
is low if compared with density estimates for other curassow species in pristine sites, and is
similar to the one reported for hunted localities.
Terborgh et al. (1990), using spot-mapping, reported a density of 5 birds/km and a
2
biomass of 15.3 kg/km for Mitu tuberosa, the razor-billed curassow, in a mature floodplain
2
forest of Amazonian Perú. Thiollay (1989) estimated the density of Crax alector, the black
curassow, in the continuous primary humid lowland rain forest of French Guiana, using strip-
transect censuses with an arbitrary detection distance of 25 metres. He calculated a density of
8.37±1.09 birds/km in an area with no hunting within $50 km. Thiollay (1989, 1994) found that
2
C. alector density was reduced to 1.38±0.87 birds/km in a forest with the nearest hunting area
2
within 3-20 km away, and to 0.39±0.34 birds/km in a regularly hunted area. This showed that
2
hunting pressure reduced the density of this curassow by approximately 70-99%, even in a
primary forest.
C. r. griscomi density (and biomass) is about 10-20% of the estimate of a curassow
density in pristine tropical forest. Considering this evidence, the low density of the Cozumel
curassow is likely to be strongly influenced by a continuous hunting pressure on the population
since the second half of the last century, when the island was again inhabited by humans.
Thiollay (1989, 1994) stated that the low curassow densities in hunted localities in French Guiana
are probably maintained only by the immigration of birds from surrounding areas that are not
subjected to hunting. C. r. griscomi population, however, is more likely to be reduced, even to
extinction, if hunting continues, because, as an isolated population, there is no "buffer effect" by
immigration, as there is on the mainland.
43
The estimated density for C. alector in the primary humid lowland forest of El Caura
Forest Reserve in Venezuela was 14.01 birds/km in June 1985, 10.16 birds/km in November
22
1985, and 7.75 birds/km in November 1986 (Silva and Strahl, 1991). These density values were
2
obtained by using mean perpendicular detection distances from the line transect (Overton, 1971).
The equivalent density for C. r. griscomi using this method was 1.7 birds/km , which also shows
2
a very low population density. The reduction in C. alector density was attributed to an increase
in subsistence hunting pressure, logging exploration and the paving of the access road in that
area. Thiollay (1992) also found that C. alector was eliminated in an area of continuous primary
lowland rain forest of French Guiana, because hunting pressure was much aggravated and
extended into the forest by logging roads, and also because of the influence of selective logging
on forest structure. These circumstances may also be reflected in the present conservation status
of C. r. griscomi. From the 1920s to the 1950s, when the latex from Manilkara zapota trees was
harvested in the forests of the island, the creation of accesses into the forest probably increased
hunting pressure on the Cozumel curassow population. By the late 1940s, this cracid was
considered almost extinct (Paynter, 1955). Additionally, the construction of the paved roads in
the 1970s furthered hunting pressure and habitat destruction, as stated by local informants, who
said that prior to the 1970s, the Cozumel curassow used to be a commoner bird on the island than
nowadays.
It has to be pointed out that the curassow density estimates reviewed are not strictly
comparable, since they were calculated using different methods. However, if the assumptions
of the method used are met, reliable estimates can be obtained. The estimates of Thiollay (1989),
are particularly comparable with mine, because he analysed his data obtained from strip-transect
census method with a detection distance of 25 metres, and I analysed mine by using a uniform
function with a strip half-width of 23.8 metres for density estimations. On the other hand,
density estimates obtained from mean perpendicular detection distances (Silva and Strahl, 1991),
are very likely to be overestimates (18-90%) of the true population density according to Robinette
et al. (1974) and Buckland et al. (1993).
44
5.2 POPULATION SIZE OF THE COZUMEL CURASSOW
The estimated population size of C. r. griscomi, was considered only for the tropical
semi-deciduous and the low tropical deciduous forests (349 km ), wherein it is likely to be
2
restricted nowadays. Although there is not sufficient recent evidence, it is possible that this
cracid occasionally uses mangrove areas and other vegetation types as well, especially those next
to the tropical semi-deciduous and low tropical deciduous forests. According to anecdotal
information, this cracid used to inhabit nearly the entire island before the 1940s, so it was not
unusual to see the Cozumel curassow in mangroves and on the beaches of the island. Island birds
tend to use a wider range of habitats than their mainland counterparts (Crowell, 1962; MacArthur
et al., 1966).
Considering that 11% of the island forests (44 km ) have been transformed because of
2
human settlements and human activities, the population size of this cracid based on the
ecological density (0.87 birds/km ) estimated should be of about 340 curassows instead of 304.
2
Hence, although marginal, there is also a reduction in the population size of this cracid, owing
to the loss of habitat only.
The density estimated for this curassow is a low density estimate for a curassow
population, as already stated. Thus, it can be roughly assumed that a relatively "healthy"
Cozumel curassow population, should have a density of about 5-8 birds/km , based on other
2
curassow density estimates in pristine habitats (Thiollay, 1989,1994; Terborgh et al., 1990; Silva
and Strahl, 1991). Consequently, the population size of the Cozumel curassow on the entire
island in pristine conditions, probably was between 1,700 and 3,400 birds. If these numbers are
considered as the C. r. griscomi population size that existed just before the human resettlement
on the island, the present estimated population size of this endemic cracid, probably represents
a reduction of about 90%, since the second half of the last century.
The estimated values of the instantaneous rate of population increase (r=-0.11) and the
finite rate of population increase (8=0.90) indicate a decline in the Cozumel curassow population
under the specific environmental conditions present at the time when the study was carried out.
Likewise, if the assumptions for the estimation of these parameters are correct, the net
0
reproductive rate (R) shows that 44% of the population is lost each generation (5.5 years).
However, this information is only preliminary and must be considered with care. For instance,
45
if the Cozumel curassow is considered to have a life-span, not of 7 years, but more than 10 years,
all the parameters show a slight increase in population size (see Appendix II). Furthermore,
because environmental conditions are not stable, particularly in an island, the value of r is
expected to change as a consequence of changes in birth and death rate. In conclusion, studies
to obtain more reliable estimates of these parameters are needed.
5.3 HABITAT PREFERENCE
5.3.1 Cenotes Influence on the Cozumel Curassow Abundance and Distribution
Cenotes and aguadas can be considered as keystone water sources for wildlife throughout
the dry season in Cozumel Island, and probably in most of the Yucatán Peninsula, since they
represent the only source of freshwater available during that period.
Based on my results, the area within 2 km radial distance from cenotes and aguadas can
be considered of high importance for the conservation of the Cozumel curassow, since the
density of this cracid is higher within this area, than beyond that radial distance from cenotes and
aguadas, at least during the dry season. The clustering of C. r. griscomi at the first 250 metres
interval of radial distance from cenotes and aguadas, is difficult to explain with the current
information available, but probably the curassows whose territories are located near cenotes and
aguadas, cluster around these water sources during the dry season. Likewise, the reason why
females cluster more than males around cenotes and aguadas, still requires more research.
The conservation of C. r. griscomi is certainly closely linked with the conservation of the
cenotes and aguadas in the island, which are of critical importance. However, because local
hunters empirically know the importance of these water sources for the Cozumel curassow (and
other wildlife species), this cracid is more frequently hunted near the cenotes and aguadas during
the dry season. Consequently, it is essential to protect these areas from hunting, since it would
diminish hunting pressure on the Cozumel curassow significantly.
Additionally, since cenotes are fed by an underground freshwater layer that lies on a salt
water layer, and the water wells that provide water for the human population of the island also
use that freshwater layer; the overexploitation of water wells for human use, can threat the water
46
quality and water level of cenotes. Thus, a management program which ensures a sustained
water quality and water level in cenotes is paramount.
5.3.2 Human Influence on the Cozumel Curassow Abundance and Distribution
The following bands of human influence on the Cozumel curassow occurrence and
density were detected:
1) 0-1.5 km from human settlements and accesses, where C. r. griscomi is absent.
2) 1.5-3 km from human settlements and accesses, where occurrence is low and, therefore,
density is depleted to about 0.11 to 0.46 birds/km .
2
3) 3-4.5 km from human settlements and accesses, where occurrence is higher than in the
previous interval, in part because of the presence of cenotes, increasing density to 0.46
to 0.77 birds/km (Figure 4.5).
2
4) >4.5 km from human settlements and accesses, where human influence is at its minimum in
the island, and density tends to achieve the estimated density for the whole population
(0.87 birds/km ).
2
These changes on C. r. griscomi density with regard to human influence agree with those
on C. alector reported by Thiollay (1989) due to hunting (see section 5.1). Thus, these findings
suggest that human influence (and probably hunting specifically) have had an important adverse
effect on the Cozumel curassow population, and may account, at least partially, for the actual
conservation status of the subspecies.
The results may also account for the low C. r. griscomi density detected in the low
tropical deciduous forest, since most of this forest type is located within 0 to 3 km from human
settlements and roads.
5.4 HURRICANES AND WILDFIRES
Studies of the effects of hurricanes on bird communities of forests in the Caribbean region
indicate that species inhabiting the mature forest interior, and nectarivorous and frugivorous bird
species decline in the months following a hurricane. These species are more severely affected
than generalist bird species, and those that primarily depend on foods other than nectar and fruit
47
(Askins and Ewert, 1991; Lynch, 1991; Waide, 1991a; Will, 1991; Wunderle, et al., 1992).
Although C. r. griscomi probably always has been subjected to the effects of hurricanes, there
was an apparently healthy population at the end of the last century, from the reports of the first
collectors who visited Cozumel Island (Nelson, 1926). The effects of hurricanes on the Cozumel
curassow population, depend on their intensity and frequency. Although there is not information
on the effect of hurricanes on the population of this cracid, based on the information available
on the effect of hurricanes on other communities in the Caribbean region, the following scenarios
are possible:
1) It is likely that low-intensity hurricanes, tropical storms and tropical depressions, have more
beneficial than adverse effects on the Cozumel curassow population and its habitat. This is
because of the high productivity brought about in the forest as a result of a sudden increase in
nutrient availability from the abundant litterfall generated and the huge quantities of water that
fall in a short period of time (Lugo et al., 1983; Frangi and Lugo, 1991; Lodge et al., 1991;
Whigham et al., 1991). Consequently, it is likely that in the years following a low-intensity
hurricane, there may be a greater production of forest fruit (Lugo et al., 1983), which would
favour the Cozumel curassow population. Thus, it is likely that the C. r. griscomi population had
not suffered important reductions because of the effects of hurricanes during the first half of this
century, since only low intensity hurricanes occurred at that time. Additionally, hurricanes can
temporarily reduce hunting pressure on the population of this cracid, because access to the forest
interior is significantly reduced, due to the impenetrable tangle produced in the forest.
2) When a strong hurricane strikes the island, some curassows may be injured or killed as a result
of the projectiles thrown by the wind. The curassows that survived the strike of the hurricane
could then be subjected to famine because of the decrease in fruit production or even lack of fruit
at all (Brokaw and Walker, 1991; Waide, 1991b; Whigham et al., 1991; You and Petty, 1991).
They may be also exposed to thirst, since the cenotes and aguadas, the only water sources during
the dry season, can be contaminated with salt water (e.g. Blood et al., 1991; local informants,
personal communication). It is possible that this situation makes the Cozumel curassow prone
to diseases and parasites, and further deaths could occur, although there is not yet evidence of
this. Additionally, because of famine, the Cozumel curassow tends to move more in search of
48
food, making itself more vulnerable to hunting or killing when it approaches settlements, paths
or roads (Paynter, 1955; local informants, personal communication). Due to the possible
reduction in population size after a strong hurricane, subsequent hunting or killing after the strick
of a hurricane can be even more serious.
Strong hurricanes are likely to reduce the Cozumel curassow population according to the
intensity of the cyclone. Unfortunately, quantitative data are still not available. Hurricane
Gilbert (September, 1988) probably caused a reduction in the C. r. griscomi population, thus its
low density and population size estimated in this study is likely to still be the consequence of this
unusually severe cyclone and the hunting that still exists in Cozumel Island. Additionally,
Hurricane Roxanne (October, 1995) might have caused a further reduction in the population size
of this cracid. Hurricanes are unlikely, however, to be responsible for its conservation status,
since Paynter (1955) considered this cracid "on the very brink of extinction" when he visited the
island on January 1949, before the strongest hurricanes of this century struck Cozumel Island.
Wildfires are associated with the dry season (Pérez-Villegas, 1980) and also with
hurricanes, which make the area affected prone to wildfires during the following dry season.
Wildfires appear to have a more severe and long-lasting impact on tree mortality, forest structure
and wildlife communities than the direct hurricane wind damage (Lynch, 1991; Whigham et al.,
1991). Wildfires are uncommon events in Cozumel Island and they have not been as large as
ones that have affected the mainland of the Yucatán Peninsula. The effect of highly humid
winds, that constantly influence the island, may account for the low occurrence of wildfires.
Nevertheless, it is evident that the risk of important wildfires exists, which represents a potential
catastrophic event that may affect considerably the survival of C. r. griscomi, even more than a
hurricane.
5.5 HUNTING
Hunting has probably been the most important human-induced cause of reduction in the
Cozumel curassow population since the human resettlement of the island. Considering the
evidence of the historical hunting intensity and the estimated present density of this cracid, in the
49
light of other cracid density estimates, the current conservation status of C. r. griscomi seems to
be mainly the consequence of hunting, particularly before the 1950s. The hunting tradition of
the inhabitants of Cozumel Island, however, has decreased in the last decades, because of the
reduction of some species of game fauna and the change in the economic activities towards
tourism. The hunters of the island are elders who are unlikely to be replaced by younger people
from Cozumel Island, as also noted by Jorgenson (1993, 1995) in a locality in the mainland of
the Yucatán Peninsula. Nevertheless, when people from the mainland come to work on the
ranches of the island, if they are used to hunt, they may hunt when they have the possibility.
Although it is quite improbable that hunting will disappear from the island, it will not achieve
the levels of previous decades in the short term. On the other hand, if the game species, and
specifically the Cozumel curassow, were able to increase their populations, it would not be
surprising if hunting increased as well, unless it was effectively banned.
5.6 PRESENT HABITAT DISTURBANCE
Human population growth and tourist development are nowadays the main threat to the
natural vegetation of the island, representing the conversion of otherwise natural areas into urban
or rural settlements and tourist resorts. Additionally, the human population growth increases the
demand for goods and services, which means pressure upon the natural resources of the island.
The principal reason for human population increase in Cozumel Island has been its quick tourist
development, since this activity requires infrastructure development and services.
It is true that agriculture and cattle ranching, that once were the main causes of natural
vegetation destruction, have decreased significantly because of the shift of economic activities
towards tourism, but the increase in human population might increase the number of people
developing these activities, bringing about the conversion of forested areas and a delay in the
succession of secondary vegetation to primary forest.
Although more than 75% of Cozumel Island is still forested and a limit for the growth
of the city of San Miguel, as well as for the human population, have already been settled, these
aims have to be accomplished effectively in order to avoid any negative effect on the biological
diversity of the island because of natural habitat lost and overexploitation.
50
5.7 THE BOA AN INTRODUCED PREDATOR
Alien faunal species introduced intentionally or unintentionally by people can exterminate
native species by competing with them, preying upon them or destroying their habitat, and their
effects are generally greatest on islands (e.g. Savidge, 1987).
Current information is not enough to state whether or not the introduction of the boa to
Cozumel Island might account for the reduction of C. r. griscomi population. Although there is
no direct evidence, the boa can be a predator of the Cozumel curassow. Boas of about 1.5 metres
long are likely to prey on eggs and chicks younger than six months, but boas of about 2.5 metres,
although not common, could kill adult-sized curassows as well (e.g., Janzen, 1970; Chapman,
1986).
Almost 30 years after its introduction, the boa is widespread on the island. Its present
broad distribution may be explained by its high reproductive rate and the lack of predators. The
Cozumel Island raccoon (Procyon pygmaeus), the coati (Nasua narica nelsoni) and the roadside
hawk (Buteo magnirostris gracilis) can be potential predators of small boas (50 cm), but big boas
have no predators, except for humans. Moreover, the island raccoon and the coati are found in
very low numbers nowadays (personal observation), and they can even be preyed on by large
boas (Janzen, 1970).
The boa can be a serious problem for the conservation of endemic fauna, especially small-
sized animals that carry out their activities mainly in the understorey of the forest, e.g. the
Cozumel thrasher (Toxostoma guttatum), endemic rodents (Reinthrodontomys spectabilis,
Oryzomys couesi cozumelae and Peromyscus leucopus cozumelae) and lizards (Cnemidophorus
cozumela cozumela), besides the Cozumel curassow. Consequently, whether or not the boa
represents a clear and present threat for C. r. griscomi, it is a potential threat for the biological
diversity of the island.
5.8 CONSERVATION STATUS OF THE COZUMEL CURASSOW
According to the results presented in this study, the conservation status of C. r. griscomi
is effectively defined as "Critically Endangered", because the population of this curassow
51
fulfilled one of the criteria proposed by the IUCN (1994) to be considered within such category:
The estimated population size held around 250 (137-468) adult curassows, which are at high risk
of being easily reduced because all the population, consisting of about 300 (164-562) birds, is
located only on Cozumel Island (criterion C2b). Additionally, the population probably has been
reduced in 90% since the mid-1800s, and it may have suffered some further reduction after the
completion of this study due to Hurricane Roxanne in the island.
The estimated population size of C. r. griscomi places this endemic curassow within one
of the smallest cracid populations, together with eleven other cracid taxa also considered as
critically endangered (Collar et al., 1992; Collar et al., 1994; Strahl et al., 1995; Galetti et al.,
in press): Ortalis vetula deschauenseei, Penelope perspicax, P. albipennis, P. jacucaca (north-
eastern Brasil population), Pipile pipile, P. jacutinga (south-eastern Brasil population),
Oreophasis derbianus, Crax alberti, C. fasciolata pinima (north-eastern Brasil population), C.
globulosa (eastern Perú and northern Bolivia population) and C. blumenbachii.
The present conservation status of this cracid is the result of a multifactorial pressure of
adverse factors over the population, as well as some intrinsic characteristics of this bird,
specifically the low reproductive rate. Furthermore, like many other insular taxa, it faces a
greater extinction probability than its mainland counterpart (C. r. rubra), because of the lack of
immigration.
Although catastrophes like hurricanes are beyond control for the conservation of the
Cozumel curassow, some other actions can be carried out to enhance its population and its
habitat, and to reduce its risk of extinction. Immediate action is required to promote the
conservation of the Cozumel curassow.
52
6
CONSERVATION OF THE
COZUMEL CURASSOW
6.1 RECOMMENDED ACTIONS FOR THE CONSERVATION OF THE
COZUMEL CURASSOW
Because extinction is a probabilistic event, and small populations are more likely to go
extinct, time is crucial for the conservation of the C. r. griscomi population. According to the
declining-population paradigm expressed by Caughley (1994), the agents responsible for the
decline in the population of C. r. griscomi have to be identified and halted to reverse such
decline. The estimation of population density and size of the Cozumel curassow can be useful
guidelines to identify and test these adverse factors and make appropriate conservation
recommendations.
6.1.1 Hunting Banning Enforcement
The evidence presented in this study shows that hunting is very likely to be one of the
main reasons of C. r. griscomi population depletion. Although hunting is legally prohibited in
Cozumel Island, this activity is still been carried out by local hunters. A fundamental step for
the conservation of the Cozumel curassow is the effective banning of hunting in the island.
Since hunter population is small nowadays in the island, activities oriented to control hunting
would be successful if enough resources (human, financial and material) are allocated, but for
an effective hunting ban, participation of the local community is essential. It is important that
hunters and local consumers understand the impact of their activities on the Cozumel curassow
population. This participation could be significantly enhanced through environmental education
activities (see below), thus former hunters and people living in rural areas and ranches could
become rangers instead of hunters.
53
6.1.2 Natural Protected Areas and Habitat Enhancement
Because forest disturbance is closely related to human population growth nowadays, it
is essential to define the carrying capacity of Cozumel Island and to set effectively a limit for
human population size on the island, for the number and size of human settlements and resorts,
for the land-use changes and a limit on the construction of roads. It means that an adequate land-
use plan for Cozumel Island has to be elaborated and effectively applied, where natural protected
areas are well defined and decreed, and also effectively protected. Natural protected areas to be
suitable for C. r. griscomi, must cover the oldest or less disturbed forested areas of the island,
because this forest seems to be the most suitable habitat for C. r. griscomi. Also these areas must
include the largest possible number of cenotes and aguadas, which have shown to be particularly
important for the Cozumel curassow. The definition of natural protected areas in the island,
however, must consider the requirements of other endemic or important wildlife species as well.
Habitat for C. r. griscomi and other wildlife species can be enhanced by creating artificial
ponds to provide water, especially during the dry season, provided these ponds are not used for
hunting. This action could be feasible near the water wells located in the forest under the
Commission for Drinkable Water and Sewerage administration, where the infrastructure already
exists.
The continuity of the island forests has been partially interrupted because of the roads
and the human settlements along the roads. However, the road segments bordered by forest are
used to a certain degree by the Cozumel curassow (and other wildlife species) for dispersion
between forest fragments. Consequently, it is essential to preserve and, if possible, expand the
connections between fragments, to avoid a further reduction in population size because of the
split of the population, which would increase the risk of extinction. It is also important to have
a strict regulation on the speed limit of vehicles using the roads, as well as to set signs along the
roads, showing the possibility of wildlife crossing the road, specially in the transversal road and
in the southern portion of the coastal road, near the mangrove area.
6.1.3 Eradication of Exotic Fauna
Exotic fauna already occurring on Cozumel Island must be completely eradicated to stop
their adverse effects on the biological diversity of the island. Likewise, domestic dogs and cats
must be prevented from becoming feral.
54
With the available information, it is not possible to conclude whether the boa represents
a threat for the conservation of the Cozumel curassow, although it is clear that it is an alien
potential predator. On the other hand, the boa is likely to be a serious problem for the
conservation of the small-sized endemic fauna. Thus, eradication actions must be carried out.
The introduction of any new exotic species must be avoided, but it is important to stress
the risk of introducing C. r. rubra into the island. Because hybridization between both curassow
subspecies is possible, it would greatly affect the genetic pool of the Cozumel curassow, diluting
the differences between the subspecies. Thus, Cracidae diversity would be depleted, and an
evolutionary process of diversification would be abruptly stopped. Additionally, diseases and
parasites affecting C. r. rubra in the mainland can infect C. r. griscomi, if the first were
introduced into Cozumel Island, representing a further pressure upon the endemic population.
Consequently, no attempts should be made to introduce C. r. rubra into Cozumel Island.
6.1.4 Environmental Education
Participation of the local community is essential for the success of the conservation
actions. This participation may be enhanced through environmental education activities,
highlighting the importance of the conservation of the Cozumel curassow and the endemic biota
of the island and their natural habitat, as well as the need to stop hunting and to eradicate the
exotic fauna. These activities have been carried out efficiently by Museo de la Isla de Cozumel
(Fundación de Parques y Museos de Cozumel), but a more active participation of the community
is needed.
6.1.5 Captive Breeding and Reintroduction Programme of the Cozumel Curassow
Captive breeding of C. r. griscomi and its subsequent reintroduction to the island forests,
should be used as an important and viable tool to increase the wild population. In captivity, the
reproductive rate of the Cozumel curassow can be increased by inducing females to produce a
second clutch by removing the first one (Ollson, 1981; Quinto-Adrián, 1981; Valenzuela, 1981).
A reintroduction programme can be viable, since there is still enough forest left that could
support a bigger population, although for this programme to be successful, it must be
accompanied by an effective enforcement of the hunting ban and an active environmental
education programme.
55
Probably, a good place to locate the captive breeding station in Cozumel Island would be
within the installations of the Commission for Drinkable Water and Sewerage in the area of water
wells of the Ejes 6 and 8, because the infrastructure (water and electricity) could be used. In
addition, the station would be placed practically in the centre of the tropical semi-deciduous
forest of the island, and human influence could be controlled in the area, making easier the
reintroduction process. However, special care must be taken with the health of captive
curassows, since a disease outbreak could spread easier into the wild population, if the breeding
station is located within the forest. Besides, the captive breeding station at Cozumel Island, it
would be necessary to establish a network of captive breeding stations outside the island in
México and other countries, to reduce the risk of extinction owing to catastrophes like unusually
strong hurricanes and wildfires. However, the Cozumel curassows exported to other breeding
stations, as well as their descendants should be property of Cozumel Island, México. These
breeding stations outside Cozumel Island must avoid any risk of hybridization between C. r.
griscomi and C. r. rubra, and a network of exchange of the captive C. r. griscomi genetic pool
has to be established to avoid inbreeding depression and genetic drift. Additionally, these
breeding stations would be important tools to support environmental education programs within
and outside Cozumel Island.
Important basic information for the management of the Cozumel curassow could be
obtained from captive birds. Additionally, the reintroduction of captive-reared Cozumel
curassows, can be an important tool to study this cracid in wild, for instance, by radio-tagging
the reintroduced birds.
6.2 RESEARCH NEEDS
The available information on the biology of C. r. griscomi is inadequate, since this cracid
had never been systematically studied in the field. Research is needed to know the population
trends and the population structure of this cracid, thus a monitoring programme is required to
exhibit changes in population size, density and structure. The population tendencies observed
have to be correlated with prevalent adverse factors, as well as with those actions carried out in
56
order to enhance the population. This may show the magnitude of the effect of adverse factors,
as well as the effectiveness of the actions to enhance C. r. griscomi population.
Radio-telemetry studies would provide valuable information about home range,
movements, habitat use and survival, which represent basic information for the management of
the population. Additionally, studies about food habits, breeding, behaviour, predators and
parasites and diseases would contribute relevant data for population management.
Studies about changes in land-use and natural vegetation cover on the island, also would
be of paramount importance, since these facts have significant implications for the conservation
of biological diversity.
Finally, the study of the ecology of the Cozumel curassow is an excellent case study to
improve our knowledge about conservation of insular biota.
57
LITERATURE CITED
Antochiw, M. and A. C. Dachary. 1991. Historia de Cozumel. Consejo Nacional para la
Cultura y las Artes. México. 413 pp.
Arnaiz-Burne, S. M. 1988. Cozumel: Del repoblamiento al auge turístico. pp. 83-88. In:
Memorias del I Encuentro de historia sobre la Isla de Cozumel. Cozumel: Un encuentro
en la historia. Cozumel, México.
Askins, R. A. and D. N. Ewert. 1991. Impact of Hurricane Hugo on bird populations on St.
John, U.S. Virgin Islands. Biotropica. 23(4a):481-487.
Blake, E. R. 1977. Manual of Neotropical birds. Volume 1. Sphenisidae (Penguins) to Laridae
(Gulls and Allies). The University of Chicago Press. USA. 674 pp.
Blood, E. R., P. Anderson, P. A. Smith, C. Nybro, and K. A. Ginsberg. 1991. Effects of
Hurricane Hugo on coastal soil solution chemistry in South Carolina. Biotropica.
23(4a):348-355.
Brokaw N. V. L. and L. R. Walker. 1991. Summary of the effects of Caribbean hurricanes on
vegetation. Biotropica. 23(4a):442-447.
Buchholz, R. 1991. Older males have bigger knobs: Correlates of ornamentation in two species
of curassow. The Auk. 108:153-160.
Buckland, S. T., D. R. Anderson, K. P. Burnham and J. L. Laake. 1993. Distance sampling:
Estimating abundance of biological populations. Chapman and Hall. London, UK. 446
pp.
Butterlin, J. and F. Bonet. 1963. Mapas geológicos de la Península de Yucatán. I.-Las
formaciones cenozoicas de la parte mexicana de la Península de Yucatán. Ingeniería
Hidráulica en México. 17(1):63-71.
Cabrera-Cano, E., M. Sousa-Sánchez, O. Téllez-Valdés and A. López-Ornat. 1982. Imágenes
de la flora quintanarroense. Centro de Investigaciones de Quintana Roo. México. 224
pp.
Castro-Sariñana, M. A. 1988. Cozumel hoy y mañana. pp. 89-104. In: Memorias del I
Encuentro de historia sobre la Isla de Cozumel. Cozumel: Un encuentro en la historia.
Cozumel, México.
58
Caughley, G. 1994. Directions in conservation biology. Journal of Animal Ecology. 63(2):215-
244.
Chapman, C. A. 1986. Boa constrictor predation and group response in white-faced cebus
monkeys. Biotropica. 18(2):171-172.
Coblentz, B. E. 1990. Exotic organisms: A dilema for conservation biology. Conservation
Biology. 4(3):261-265.
Collar, N. J., L. P. Gonzaga, N. Krabbe, A. Madroño Nieto, L. G. Naranjo, T. A. Parker III and
D. C. Wege. 1992. Threatened birds of the Americas. The ICBP/IUCN Red Data Book.
3 edition. International Council for Bird Preservation. Cambridge, UK. 1150 pp.
rd
__________, M. J. Crosby and A. J. Stattersfield. 1994. Birds to watch 2. The world list of
threatened birds. BirdLife Conservation Series No. 4. BirdLife International. Great
Britain. 407 pp.
Crow, T. R. 1980. A rainforest chronicle: A 30-year record of change in structure and
composition at El Verde, Puerto Rico. Biotropica. 12(1):42-55.
Crowell, K. L. 1962. Reduced interspecific competition among the birds of Bermuda. Ecology.
42:75-88.
Cuarón, A. D. 1991. Conservación de los primates y sus hábitats en el sur de México. Tesis de
Maestría. Universidad Nacional. Heredia, Costa Rica. 113 pp.
Dachary, A. C. and S. M. Arnaiz-Burne. 1988. Repoblamiento y desarrollo de la Isla de
Cozumel 1847-1900. pp. 69-81. In: Memorias del I Encuentro de historia sobre la Isla
de Cozumel. Cozumel: Un encuentro en la historia. Cozumel, México.
Davidson, W. V. 1975. The geographical setting. pp. 47-59. In: J. A. Sabloff and W. L. Rathje
(eds.). Changing pre-Columbian commercial systems. Peabody Museum Monograph
No. 3. Harvard University. USA.
Decker, D. M. 1991. Systematics of the coatis, genus Nasua (Mammalia:Procyonidae).
Proceedings of the Biological Society of Washington. 104(2):370-386.
Delacour, J. and D. Amadon. 1973. Curassows and related birds. American Museum of Natural
History. New York, USA. 247 pp.
Ferguson-Lee, J. and E. Faull. 1992. Endangered birds. George Philip Ltd. London, UK. pp.
42, 45.
59
Frangi, J. L. and A. E. Lugo. 1991. Hurricane damage to a flood plain forest in the Luquillo
Mountains of Puerto Rico. Biotropica. 23(4a):324-335.
Freidel, D. A. and J. A. Sabloff. 1984. Cozumel: Late Maya settlement patterns. Academic
Press, Inc. USA. 208 pp.
Galetti, M., P. Martuscelli, F. Olmos and A. Aleixo. In press. Ecology and conservation of the
jacutinga Pipile jacutinga in the Atlantic forest of Brazil. Biological Conservation.
Guzmán-Grajales, S. M. and L. R. Walker. 1991. Differential seeding responses to litter after
Hurricane Hugo in the Luquillo Experimental Forest, Puerto Rico. Biotropica.
23(4a):407-413.
Hall, E. R. 1981. The mammals of North America. Vols. I-II. 2 edition. Wiley Interscience
nd
Publication. USA. 1181 pp. + index.
Hamblin, N. L. 1984. Animal use by the Cozumel Maya. The University of Arizona Press.
Arizona, U.S.A. 206 pp.
__________ and A. M. Rea. 1979. La avifauna arqueológica de Cozumel. Boletín de la
Escuela de Ciencias Antropológicas de la Universidad de Yucatán. 37(7):21-49.
Howell, S. N. G. and S. Webb. 1995. A guide to the birds of Mexico and northern Central
America. Oxford University Press. New York, USA. 851 pp.
Hoyo, J. del, A. Elliot and J. Sargatal (eds.). 1994. Handbook of the birds of the world. Vol 2:
New world vultures to Guinea fowl. Lynx Editions. Barcelona, Spain.
INEGI. 1994. Cozumel, Estado de Quintana Roo. Cuaderno Estadístico Municipal. Edición
1993. Instituto Nacional de Estadística, Geografía e Informática. México. 117 pp.
IUCN Species Survival Commission. 1994. IUCN Red list categories. IUCN. Gland,
Switzerland. 21 pp.
Janzen, D. H. 1970. Altruism by coatis in the face of predation by Boa constrictor. Journal of
Mammalogy. 51(2):387-389.
Jáuregui, E., J. Vidal and F. Cruz. 1980. Los ciclones y tormentas tropicales en Quintana Roo
durante el periodo 1871-1978. pp. 47-63. In: Memorias de problemática y perspectiva
de Quintana Roo. Instituto de Geografía, UNAM y CIQRO. México.
Jorgenson, J. P. 1993. Gardens, wildlife densities and subsistence hunting by Maya indians in
Quintana Roo, México. PhD Thesis. University of Florida. Gainesville, Florida, USA.
336 pp.
60
__________. 1995. Maya subsistence hunters in Quintana Roo, México. Oryx. 29(1):49-57.
Laake, J. L., S. T. Buckland, D. R. Anderson, and K. P. Burnham. 1994. Distance user's guide
V2.1. Colorado Cooperative Fish and Wildlife Research Unit Colorado State University.
Fort Collins, CO, USA. 84pp.
Landsea, C. 1996. The North Atlantic "best track": 1886-1996. National Hurricane Center.
Miami, Florida, USA. http://wxp.atms.purdue.edu/hur-atlantic/
Lee, J. C. 1996. The amphibians and reptiles of the Yucatán Peninsula. Cornell University
Press. New York, USA. 500 pp.
Lesser-Jones, H., J. Azpetia-Reyes, and J. M. Lesser-Illades. 1977. Geohidrología de la Isla de
Cozumel, Quintana Roo. Proyecto CONACyT NSF 704 SARH. Unpublished
manuscript.
Lodge, D. J., F. N. Scatena, C. E. Asbury and M. J. Sánchez. 1991. Fine litterfall and related
nutrient inputs resulting from Hurricane Hugo in subtropical and lower montane rain
forest of Puerto Rico. Biotropica. 23(4a):336-342.
Lugo, A. E., M. Applefield, D. J. Pool, and R. B. McDonald. 1983. The impact of Hurricane
David on the forests of Dominica. Canadian Journal of Forestry Research. 13:201-211.
Lynch, J. F. 1991. Effects of Hurricane Gilbert on birds in a dry tropical forest in the Yucatán
Peninsula. Biotropica. 23(4a):488-496.
MacArthur, R. H., H. Recher and M. Cody. 1966. On the relation between habitat selection and
species diversity. American Naturalist. 100:319-332.
Matola, S. M., A. D. Cuarón and H. Rubio-Torrler. In press. Status and action plan of Baird's
tapir (Tapirus bairdii). In: D. M. Brooks, R. E. Bodmer and S. M. Matola (eds.). Tapirs:
Status survey and conservation action plan. IUCN. World Conservation Union. Species
Survival Commission.
Merino-Ibarra, M. and L. Otero-Davalos. 1991. Atlas ambiental costero. Instituto de Ciencias
del Mar y Limnología-Centro de Investigaciones de Quintana Roo. Quintana Roo,
México. 80 pp.
Miranda, A., A. Espinosa de los Monteros, M. del C. Arizmendi, J. López-Portillo, J. P. Gallo,
and H. Berlanga. 1988. Evaluación del impacto ambiental del proyecto: Parque
arqueológico Isla de la Pasión. Cozumel, Quintana Roo. Centro de Ecología, UNAM.
México. 105 pp.
61
Monroe, B. L. and C. G. Sibley. 1993. A world checklist of birds. Yale University Press. USA.
393 pp.
Morales, J. J. 1993. Los huracanes en la Península de Yucatán. 2 edition. Talleres Gráficos
nd
del Sudeste. Mérida, Yucatán, México. 111 pp.
Nelson, E. W. 1926. Two new birds from Mexico. Proceedings of the Biological Society of
Washington. 39:105-108.
Ollson, M. 1981. Captive propagation of cracids. pp. 227-230. In: Memorias del Primer
Simposio Internacional de la Familia Cracidae. Facultad de Medicina, Veterinaria y
Zootecnia, UNAM. México.
Overton, W. S. 1971. Estimating the numbers of animals in wildlife populations. pp. 403-457.
In: R. H. Giles, Jr. (ed.). Wildlife management techniques. 3 edition. The Wildlife
rd
Society. Washington, D.C., USA. 633 pp.
Paynter, R. A., Jr. 1955. The ornithogeography of the Yucatán Peninsula. Peabody Museum of
Natural History. Bulletin 9. Yale University. New Heaven, Connecticut, USA. 347 pp.
Pérez-Villegas, G. 1980. El clima y los incendios forestales en Quintana Roo. pp. 65-80. In:
Memorias de problemática y perspectiva de Quintana Roo. Instituto de Geografía,
UNAM y CIQRO. México.
Pielou, E. C. 1974. Population and community ecology. Principles and methods. Gordon and
Breach Science Publishers. New York, USA. 424 pp.
Quinto-Adrián, J. F. 1981. Observaciones y reproducción de Crax rubra en condiciones
seminaturales en San Felipe Bacalar, Quintana Roo, México. pp. 249-259. In: Memorias
del Primer Simposio Internacional de la Familia Cracidae. Facultad de Medicina,
Veterinaria y Zootecnia, UNAM. México.
Redford, K. H. 1992. The empty forest. BioScience. 42(6):412-422.
__________ and J. G. Robinson. 1987. The game of choice: Patterns of indian and colonist
hunting in the neotropics. American Anthropologist. 89:650-667.
Richards, H. G. 1937. Land and freshwater mollusks from the Island of Cozumel, México, and
their bearing on the geological history of the region. Proceedings of the American
Philosophical Society. 77(3):249-262.
62
Ridgway, R. 1885. Catalogue of a collection of birds made on the Island of Cozumel, Yucatán,
by the naturalists of the U. S. Fish Commission Steamer Albatross, Capt. Z. L. Tanner,
Commander. Proceedings of the United States National Museum. 8(36):560-583.
__________ and H. Friedmann. 1946. Birds of North and Middle America. Bulletin of the
United States National Museum. 50(10):484 pp.
Robinette, W. L., C. M. Loveless and D. A. Jones. 1974. Field tests of strip census methods.
Journal of Wildlife Management. 38(1):81-96.
Rodríguez-Estrella, R., J. L. León de la Luz, A. Braceda, A. Castellanos, J. Cancino, and J.
Llinas. 1996. Status, density and habitat relationships of the endemic terrestrial birds of
the Socorro Islands, México. Biological Conservation. 76:195-202.
Sabloff, J. A. and W. L. Rathje (eds.). 1975. Changing pre-Columbian commercial systems.
Peabody Museum Monograph No. 3. Harvard University. USA.
Salvin, O. 1889. A list of the birds of the islands of the Coast of Yucatán and of the Bay of
Honduras. The Ibis. Sixth Series. 1:359-379.
Savidge, J. A. 1987. Extinction of an island forest avifauna by an introduced snake. Ecology.
68(3):660-668.
SEDESOL (Secretaria de Desarrollo Social). 1994. Norma Oficial Mexicana NOM-059-ECOL-
1994, que determina las especies y subespecies de flora y fauna silvestres terrestres y
acuáticas en peligro de extinción, amenazadas, raras y las sujetas a protección especial,
y que establece especificaciones para su protección. Diario Oficial de la Federación.
Organo del Gobierno Constitucional de los Estados Unidos Mexicanos. 488(10):2-60.
Sermeño-Martinez, A. 1986. Alimentación y reproducción del pajuil (Crax rubra) en El
Salvador. Tesis de Licenciatura. Universidad de El Salvador. San Salvador, El
Salvador. 61 pp.
Sick, H. 1993. Birds in Brazil. Princenton University Press. New Jersey, USA. pp. 189-201.
Silva, J. L. and S. D. Strahl. 1991. Human impact on populations of chachalacas, guans, and
curassows (Galliformes:Cracidae) in Venezuela. pp. 37-52. In: J. G. Robinson and K.
H. Redford (eds.). Neotropical wildlife use and conservation. The University of Chicago
Press. USA.
63
Smith, H. M. and E. H. Taylor. 1945. An annotated checklist and key to the snakes of México.
Smithsonian Institution, United States National Museum. Bulletin 187. Washington,
D.C., USA. 239 pp.
Smithe, F. B. 1974. Naturalist's color guide supplement. The American Museum of Natural
History. New York, USA. 229 pp.
__________. 1975. Naturalist's color guide. The American Museum of Natural History. New
York, USA.
__________. 1981. Naturalist's color guide. Part III. The American Museum of Natural
History. New York, USA. 37 pp.
Sokal, R. R. and F. J. Rohlf. 1995. Biometry. 3 edition. Freeman and Company. New York,
rd
USA. 887 pp.
SPP. 1979. Atlas de huracanes en el Océano Pacífico y en el Océano Atlántico. Secretaría de
Programación y Presupuesto. México. 49 pp.
Strahl, S. D. 1990. Conservation strategy and action plan: 1990-1995. International Cracidae
Specialist Group. WPA/ICBP. Unpublished manuscript. 72 p.
__________ and A. Grajal. 1991. Conservation of large avian frugivores and the management
of Neotropical protected areas. Oryx 25(1):50-55.
__________, S. Ellis, O. Byers and C. Plasse (eds.). 1995. Conservation assessment and
management plan for Neotropical guans, curassows and chachalacas. Working draft.
IUCN/SSC. Conservation Breeding Specialist Group. MN, USA. 183 pp + appendices.
Taylor, H. L. and C. R. Cooley. 1995. A multivariate-analysis of morphological variation
among parthenogenetic teiid lizards of the Cnemidophorus-cozumela complex.
Herpetologica. 51(1):67-76.
Téllez-Valdés, O. and E. F. Cabrera-Cano. 1987. Listados florísticos de México. VI. Florula
de la Isla de Cozumel, Q. R. Instituto de Biología, UNAM. México. 34 pp.
__________, E. F. Cabrera-Cano, E. Linares-Mazari, and R. Bye. 1989. Las plantas de
Cozumel. Instituto de Biología, UNAM. México. 75 pp.
Terborgh, J., S. K. Robinson, T. A. Parker III, C. A. Munn, and N. Pierpont. 1990. Structure
and organization of an Amazonian forest bird community. Ecological Monographs.
60(2):213-238.
64
Thiollay, J. M. 1989. Area requirements for the conservation of rain forest raptors and game
birds in French Guiana. Conservation Biology. 3(2):128-137.
__________. 1992. Influence of selective logging on bird species diversity in a Guianan Rain
Forest. Conservation Biology. 6(1):47-63.
__________. 1994. Structure, density and rarity in an Amazonian rainforest bird community.
Journal of Tropical Ecology. 10:449-481.
Valenzuela, A. J. 1981. Management of cracids in aviculture. pp. 232-246. In: Memorias del
Primer Simposio Internacional de la Familia Cracidae. Facultad de Medicina, Veterinaria
y Zootecnia, UNAM. México.
Vaurie, C. 1967. Systematic notes on the bird family Cracidae. No. 9 The genus Crax.
American Museum Novitates. 2305:1-20.
Waide, R. B. 1991a. The effect of Hurricane Hugo on bird populations in the Luquillo
Experimental Forest, Puerto Rico. Biotropica. 23(4a):475-480.
__________. 1991b. Summary of the response of animal populations to hurricanes in the
Caribbean. Biotropica. 23(4a):508-512.
Walker, L. R. 1991. Tree damage and recovery from Hurricane Hugo in Luquillo Experimental
Forest, Puerto Rico. Biotropica. 23(4a):379-385.
Whigham, D. F., I. Olmsted, E. Cabrera-Cano, and M. E. Harmon. 1991. The impact of
Hurricane Gilbert on trees, litterfall, and woody debris in a dry tropical forest in the
northeastern Yucatán Peninsula. Biotropica. 23(4a):434-441.
Will, T. 1991. Birds of a severely hurricane-damaged Atlantic Coast rain forest in Nicaragua.
Biotropica. 23(4a):497-507.
Wilson, D. E. and D. M. Reeder (eds.). 1993. Mammal species of the world: A taxonomic and
geographic reference. 2 edition. Smithsonian Institution Press. USA. 1206 pp.
nd
Wunderle, J. M., D. J. Lodge, and R. B. Waide. 1992. Short-term effects of Hurricane Gilbert
on terrestrial bird populations on Jamaica. The Auk. 109(1):148-166.
Yih, K., D. H. Boucher, J. H. Vandermeer, and N. Zamora. 1991. Recovery of the rain forest
of southeastern Nicaragua after destruction by Hurricane Joan. Biotropica. 23(2):106-
113.
65
You, C. and W. H. Petty. 1991. Effects of Hurricane Hugo on Manilkara bidentata, a primary
tree species in the Luquillo Experimental Forest of Puerto Rico. Biotropica 23(4a):400-
406.
Appendix I - 1
APPENDIX I
MORPHOMETRIC AND PLUMAGE COMPARISONS OF THE
COZUMEL CURASSOW AND THE GREAT CURASSOW
The validity of Crax rubra griscomi as a true subspecies, different from C. r. rubra, was
analysed by comparing quantitative and qualitative characteristics from museum specimens at
The Natural History Museum (British Museum), the American Museum of Natural History and
the National Museum of Natural History (Smithsonian Institution). One hundred and twenty
curassow skins were reviewed, of which 49 were from males (3 from C. r. griscomi and 46 from
C. r. rubra) and 71 from females (6 from C. r. griscomi and 65 from C. r. rubra). The skins
reviewed comprised most of the distribution of the species to account for the variability of the
species within its range. Males and females were analysed separately because of sexual
dimorphism. Wing and tail measurements were not considered for the analysis to avoid potential
error associated to abrasion and moult of feathers.
Besides differences found in the museum specimens reviewed, during field work
differences were also detected in colour of feet of C. r. griscomi females. I observed three
females with yellowish orange feet (approximately colour 153, see Smithe, 1974, 1975, 1981),
one with pinkish or flesh colour feet (approximately colour 5, see Smithe, 1974, 1975, 1981) and
one with pinkish grey feet; out of seven females observed (colour of feet of the other two females
could not be observed). Conversely, colour of feet of C. r. rubra females has been reported only
as pinkish grey (Ridgway and Friedmann, 1946; personal observations).
Appendix I - 2
TABLE I.1 Male Crax rubra measurements analysed from skin specimens at The Natural History Museum
(British Museum), the American Museum of Natural History and the National Museum of Natural History
(Smithsonian Institution). Morphometric characters of curassows from the mainland and from Cozumel Island
were compared by t-tests. Statistical differences are shown in bold.
Character
(mm) Crax rubra rubra Crax rubra griscomi t-value df
One-tailed
significance
Bill width at
lore level
0 = 21.47 (18.30-24.65)
s = 1.60
n = 45
0 = 19.13 (18.43-19.92)
s = 0.75
n = 3
2.49 46 p = 0.008
Culmen with
cere
0 = 51.74 (38.94-59.16)
s = 4.31
n = 44
0 = 48.35 (44.10-51.88)
s = 3.94
n = 3
1.32 45 p = 0.096
Culmen
without cere
0 = 29.38 (19.61-33.75)
s = 2.96
n = 44
0 = 27.69 (25.48-29.20)
s = 1.95
n = 3
0.97 45 p = 0.169
Gonys 0 = 18.48 (14.17-20.82)
s = 1.34
n = 45
0 = 17.29 (16.03-18.42)
s = 1.20
n = 3
1.50 46 p = 0.07
Tarsus length 0 = 121.65 (95.82-131.72)
s = 6.40
n = 44
0 = 108.86 (101.64-116.22)
s = 7.29
n = 3
3.33 45 p = 0.001
Appendix I - 3
TABLE I.2 Male Crax rubra measurements analysed from skin specimens at The Natural History Museum
(British Museum), the American Museum of Natural History and the National Museum of Natural History
(Smithsonian Institution). Morphometric characters of curassows from the Yucatán Peninsula and from Cozumel
Island were compared by t-tests. Statistical differences are shown in bold.
Character
(mm)
Yucatán Peninsula
Crax rubra rubra
Cozumel Island
Crax rubra griscomi t-value df
One-tailed
significance
Bill width at
lore level
0 = 21.18 (18.99-24.51)
s = 2.48
n = 4
0 = 19.13 (18.43-19.92)
s = 0.75
n = 3
1.36 5 p = 0.116
Culmen with
cere
0 = 54.65 (52.32-57.04)
s = 2.62
n = 4
0 = 48.35 (44.10-51.88)
s = 3.94
n = 3
2.57 5 p = 0.025
Culmen
without cere
0 = 29.58 (28.39-31.11)
s = 1.26
n = 4
0 = 27.69 (25.48-29.20)
s = 1.95
n = 3
1.57 5 p = 0.088
Gonys 0 = 19.09 (18.10-19.77)
s = 0.77
n = 4
0 = 17.29 (16.03-18.42)
s = 1.20
n = 3
2.45 5 p = 0.029
Tarsus
length
0 = 123.26 (116.83-127.81)
s = 4.71
n = 4
0 = 108.86 (101.64-116.22)
s = 7.29
n = 3
3.21 5 p = 0.012
Appendix I - 4
TAB LE I.3 Female Crax rubra measurements analysed from skin specimens at The Natural History Museum
(British Museum), the American Museum of Natural History and the National Museum of Natural History
(Smithsonian Institution). Morphometric characters of curassows from the mainland and from Cozumel Island
were compared by t-tests. Statistical differences are shown in bold.
Character
(mm) Crax rubra rubra Crax rubra griscomi t-value df
One-tailed
significance
Bill width at
lore level
0 = 19.94 (16.24-23.70)
s = 1.37
n = 64
0 = 18.69 (17.23-19.35)
s = 0.77
n = 6
2.22 68 p = 0.015
Culmen with
cere
0 = 46.45 (39.18-55.33)
s = 3.08
n = 62
0 = 42.48 (40.82-43.86)
s = 1.35
n = 6
3.11 66 p = 0.001
Culmen
without cere
0 = 27.25 (21.33-30.91)
s = 2.55
n = 61
0 = 24.82 (23.38-26.49)
s = 1.05
n = 6
4.51 12.21 p < 0.001
Gonys 0 = 17.18 (13.53-20.53)
s = 1.26
n = 64
0 = 15.27 (13.98-15.93)
s = 0.70
n = 6
3.65 68 p < 0.001
Tarsus length 0 = 115.04 (96.11-
124.62)
s = 5.54
n = 64
0 = 99.96 (95.32-104.27)
s = 3.10
n = 6
6.53 68 p << 0.001
White band
in crest
0 = 28.29 (6.86-55.62)
s = 10.99
n = 61
0 = 43.94 (38.74-52.51)
s = 5.03
n = 6
-6.28 10.59 p << 0.001
Appendix I - 5
TAB LE I.4 Female Crax rubra measurements analysed from skin specimens at The Natural History Museum
(British Museum), the American Museum of Natural History and the National Museum of Natural History
(Smithsonian Institution). Morphometric characters of curassows from the Yucatán Peninsula and from Cozumel
Island were compared by t-tests. In the case of heterogeneity of variances the Welch's approximate t-test was
performed. Statistical differences are shown in bold.
Character
(mm)
Yucatán Peninsula
Crax rubra rubra
Cozumel Island
Crax rubra griscomi t-value df
One-tailed
significance
Bill width at
lore level
0 = 19.89 (17.69-23.70)
s = 1.57
n = 11
0 = 18.69 (17.23-19.35)
s = 0.77
n = 6
1.74 15 p = 0.051
Culmen with
cere
0 = 46.76 (42.92-49.84)
s = 2.60
n = 11
0 = 42.48 (40.82-43.86)
s = 1.35
n = 6
4.45 15 p << 0.001
Culmen
without cere
0 = 26.86 (21.33-30.84)
s = 2.70
n = 11
0 = 24.82 (23.38-26.49)
s = 1.05
n = 6
1.76 15 p = 0.05
Gonys 0 = 17.64 (15.20-19.39)
s = 1.34
n = 11
0 = 15.27 (13.98-15.93)
s = 0.70
n = 6
4.78 15 p << 0.001
Tarsus length 0 = 114.34 (108.01-120.48)
s = 3.99
n = 11
0 = 99.96 (95.32-104.27)
s = 3.10
n = 6
7.62 15 p << 0.001
White band
in crest
0 = 33.43 (25.30-41.48)
s = 5.31
n = 9
0 = 43.94 (38.74-52.51)
s = 5.03
n = 6
-3.83 13 p = 0.001
Appendix I - 6
TAB LE I.5 Female Crax rubra plumage characteristics analysed from skin specimens at The Natural History
Museum (British Museum), the American Museum of Natural History and the National Museum of Natural
History (Smithsonian Institution). Characters were compared by the Mann-Whitney U-test corrected for ties.
Statistical differences are shown in bold.
[4]
Character t
C. r. rubra
1
n
C. r. griscomi
2
n
Two-tailed
significance
Crest feathers pattern -1.973 65 6 p = 0.048
Auricular feathers pattern -2.306 65 6 p = 0.021
Nape feathers pattern -3.030 65 6 p = 0.002
Dorsal-neck feathers pattern -3.045 65 6 p = 0.002
Lateral-neck feathers pattern -2.190 65 6 p = 0.028
Chin feathers pattern -3.652 64 6 p < 0.001
Upper-throat feathers pattern -3.682 63 6 p < 0.001
Lower-throat feathers pattern -3.893 63 6 p<<0.001
Upper-breast feathers pattern -0.105 65 6 p = 0.917
Lower-breast feathers pattern -0.179 65 6 p = 0.858
Upper-abdomen feathers pattern -1.545 64 6 p = 0.122
Lower-abdomen feathers pattern -1.128 65 6 p = 0.259
Under-tail covert feathers pattern -1.810 63 6 p = 0.070
Flank feathers pattern -1.212 65 6 p = 0.225
Back feathers pattern -1.317 65 6 p = 0.188
Rump feathers pattern -0.974 65 6 p = 0.330
Upper-tail covert feathers pattern -1.198 62 6 p = 0.231
Rectrices pattern -2.376 63 6 p = 0.017
Primaries pattern -1.149 65 6 p = 0.251
Secondaries pattern -2.382 65 6 p = 0.017
Greater-primary coverts pattern -0.293 65 6 p = 0.769
Greater-secondary coverts pattern -1.782 64 6 p = 0.075
Alula pattern -1.120 65 6 p = 0.263
Scapular feathers pattern -0.639 65 6 p = 0.523
Crural feathers pattern -1.578 64 6 p = 0.115
Appendix II - 1
APPENDIX II
LIFE AND FERTILITY TABLES OF THE COZUMEL CURASSOW
0
Estimates of the net reproductive rate (R), the instantaneous rate of population increase
(r) and the finite rate of population increase (8) of Crax rubra griscomi based on different
supposed life spans (5 to 11 and 15 years) are presented. It was assumed that 50% of female
curassows survive to age of first breeding (Silva and Strahl, 1991), and then, the survivorship is
0.7 at all ages (a liberal estimate).The fertility rate was estimated from the results of this study
and it was assumed to be the same for all adult females, independently of age (see Chapter 4).
TAB LE II.1 Life and fertility table of Crax rubra griscomi based on a supposed life span of 5 years.
Age class in
years (x)
Age-specific
x
survival rate (l)
Age-specific
x
fertility (m)
Reproductive rate of
xx xx
each age class (lm)(xl m )
0<x#1 0.5 0 0 0
1<x#2 0.5 0 0 0
2<x#3 0.5 0 0 0
3<x#4 0.7 0.2 0.14 0.56
4<x#5 0.7 0.2 0.14 0.70
0xx
Net reproductive rate = R = 3lm = 0.28 females/female/generation
xx 0
Mean length of a generation = G = 3xl m ÷ R = 4.5 years
0
Instantaneous rate of population increase = r = ln R ÷ G = -0.28 curassows/curassow/year
xx
when 3lme=1
-rx
Finite rate of population growth = 8 = e = 0.75 curassows/curassows/year
r
Appendix II - 2
TAB LE II.2 Life and fertility table of Crax rubra griscomi based on a supposed life span of 6 years.
Age class in
years (x)
Age-specific
x
survival rate (l)
Age-specific
x
fertility (m)
Reproductive rate of
xx xx
each age class (lm)(xl m )
0<x#1 0.5 0 0 0
1<x#2 0.5 0 0 0
2<x#3 0.5 0 0 0
3<x#4 0.7 0.2 0.14 0.56
4<x#5 0.7 0.2 0.14 0.70
5<x#6 0.7 0.2 0.14 0.84
0xx
Net reproductive rate = R = 3lm = 0.42 females/female/generation
xx 0
Mean length of a generation = G = 3xl m ÷ R = 5 years
0
Instantaneous rate of population increase = r = ln R ÷ G = -0.17 curassows/curassow/year
xx
when 3lme=1
-rx
Finite rate of population growth = 8 = e = 0.84 curassows/curassows/year
r
TAB LE II.3 Life and fertility table of Crax rubra griscomi based on a supposed life span of 7 years.
Age class in
years (x)
Age-specific
x
survival rate (l)
Age-specific
x
fertility (m)
Reproductive rate of
xx xx
each age class (lm)(xl m )
0<x#1 0.5 0 0 0
1<x#2 0.5 0 0 0
2<x#3 0.5 0 0 0
3<x#4 0.7 0.2 0.14 0.56
4<x#5 0.7 0.2 0.14 0.70
5<x#6 0.7 0.2 0.14 0.84
6<x#7 0.7 0.2 0.14 0.98
0xx
Net reproductive rate = R = 3lm = 0.56 females/female/generation
xx 0
Mean length of a generation = G = 3xl m ÷ R = 5.5 years
0
Instantaneous rate of population increase = r = ln R ÷ G = -0.11 curassows/curassow/year
xx
when 3lme=1
-rx
Finite rate of population growth = 8 = e = 0.90 curassows/curassows/year
r
Appendix II - 3
TAB LE II.4 Life and fertility table of Crax rubra griscomi based on a supposed life span of 8 years.
Age class in
years (x)
Age-specific
x
survival rate (l)
Age-specific
x
fertility (m)
Reproductive rate of
xx xx
each age class (lm)(xl m )
0<x#1 0.5 0 0 0
1<x#2 0.5 0 0 0
2<x#3 0.5 0 0 0
3<x#4 0.7 0.2 0.14 0.56
4<x#5 0.7 0.2 0.14 0.70
5<x#6 0.7 0.2 0.14 0.84
6<x#7 0.7 0.2 0.14 0.98
7<x#8 0.7 0.2 0.14 1.12
0xx
Net reproductive rate = R = 3lm = 0.7 females/female/generation
xx 0
Mean length of a generation = G = 3xl m ÷ R = 6 years
0
Instantaneous rate of population increase = r = ln R ÷ G = -0.06 curassows/curassow/year
xx
when 3lme=1
-rx
Finite rate of population growth = 8 = e = 0.94 curassows/curassows/year
r
Appendix II - 4
TAB LE II.5 Life and fertility table of Crax rubra griscomi based on a supposed life span of 9 years.
Age class in
years (x)
Age-specific
x
survival rate (l)
Age-specific
x
fertility (m)
Reproductive rate of
xx xx
each age class (lm)(xl m )
0<x#1 0.5 0 0 0
1<x#2 0.5 0 0 0
2<x#3 0.5 0 0 0
3<x#4 0.7 0.2 0.14 0.56
4<x#5 0.7 0.2 0.14 0.70
5<x#6 0.7 0.2 0.14 0.84
6<x#7 0.7 0.2 0.14 0.98
7<x#8 0.7 0.2 0.14 1.12
8<x#9 0.7 0.2 0.14 1.26
0xx
Net reproductive rate = R = 3lm = 0.84 females/female/generation
xx 0
Mean length of a generation = G = 3xl m ÷ R = 6.5 years
0
Instantaneous rate of population increase = r = ln R ÷ G = -0.03 curassows/curassow/year
xx
when 3lme=1
-rx
Finite rate of population growth = 8 = e = 0.97 curassows/curassows/year
r
Appendix II - 5
TAB LE II.6 Life and fertility table of Crax rubra griscomi based on a supposed life span of 10 years.
Age class in
years (x)
Age-specific
x
survival rate (l)
Age-specific
x
fertility (m)
Reproductive rate of
xx xx
each age class (lm)(xl m )
0<x#1 0.5 0 0 0
1<x#2 0.5 0 0 0
2<x#3 0.5 0 0 0
3<x#4 0.7 0.2 0.14 0.56
4<x#5 0.7 0.2 0.14 0.70
5<x#6 0.7 0.2 0.14 0.84
6<x#7 0.7 0.2 0.14 0.98
7<x#8 0.7 0.2 0.14 1.12
8<x#9 0.7 0.2 0.14 1.26
9<x#10 0.7 0.2 0.14 1.40
0xx
Net reproductive rate = R = 3lm = 0.98 females/female/generation
xx 0
Mean length of a generation = G = 3xl m ÷ R = 7 years
0
Instantaneous rate of population increase = r = ln R ÷ G = -0.003 curassows/curassow/year
xx
when 3lme=1
-rx
Finite rate of population growth = 8 = e = 0.997 curassows/curassows/year
r
Appendix II - 6
TAB LE II.7 Life and fertility table of Crax rubra griscomi based on a supposed life span of 11 years.
Age class in
years (x)
Age-specific
x
survival rate (l)
Age-specific
x
fertility (m)
Reproductive rate of
xx xx
each age class (lm)(xl m )
0<x#1 0.5 0 0 0
1<x#2 0.5 0 0 0
2<x#3 0.5 0 0 0
3<x#4 0.7 0.2 0.14 0.56
4<x#5 0.7 0.2 0.14 0.70
5<x#6 0.7 0.2 0.14 0.84
6<x#7 0.7 0.2 0.14 0.98
7<x#8 0.7 0.2 0.14 1.12
8<x#9 0.7 0.2 0.14 1.26
9<x#10 0.7 0.2 0.14 1.40
10<x#11 0.7 0.2 0.14 1.54
0xx
Net reproductive rate = R = 3lm = 1.12 females/female/generation
xx 0
Mean length of a generation = G = 3xl m ÷ R = 7.5 years
0
Instantaneous rate of population increase = r = ln R ÷ G = 0.02 curassows/curassow/year
xx
when 3lme=1
-rx
Finite rate of population growth = 8 = e = 1.02 curassows/curassows/year
r
Appendix II - 7
TAB LE II.8 Life and fertility table of Crax rubra griscomi based on a supposed life span of 15 years.
Age class in
years (x)
Age-specific
x
survival rate (l)
Age-specific
x
fertility (m)
Reproductive rate of
xx xx
each age class (lm)(xl m )
0<x#1 0.5 0 0 0
1<x#2 0.5 0 0 0
2<x#3 0.5 0 0 0
3<x#4 0.7 0.2 0.14 0.56
4<x#5 0.7 0.2 0.14 0.70
5<x#6 0.7 0.2 0.14 0.84
6<x#7 0.7 0.2 0.14 0.98
7<x#8 0.7 0.2 0.14 1.12
8<x#9 0.7 0.2 0.14 1.26
9<x#10 0.7 0.2 0.14 1.40
10<x#11 0.7 0.2 0.14 1.54
11<x#12 0.7 0.2 0.14 1.68
12<x#13 0.7 0.2 0.14 1.82
13<x#14 0.7 0.2 0.14 1.96
14<x#15 0.7 0.2 0.14 2.1
0xx
Net reproductive rate = R = 3lm = 1.68 females/female/generation
xx 0
Mean length of a generation = G = 3xl m ÷ R = 9.5 years
0
Instantaneous rate of population increase = r = ln R ÷ G = 0.05 curassows/curassow/year
xx
when 3lme=1
-rx
Finite rate of population growth = 8 = e = 1.06 curassows/curassows/year
r
Appendix III - 1
APPENDIX III
CENOTES INFLUENCE ON THE COZUMEL CURASSOW DENSITY
TAB LE III.1 Crax rubra griscomi cumulative density estimates for the areas within different radial distances
from cenotes and aguadas (Chapter 4, Figure 4.1).
Radial
distance from
cenote (km)
Cumulative
density estimate
(birds/km )
2
Confidence
Interval
(95%)
Standard
Error
% Coef.
Variation
Sample size
(transects
surveyed)
Sampling
effort (km)
0.25 5.25 2.84 - 9.69 1.64 31.29 51 36.03
0.50 3.34 1.78 - 6.29 1.08 32.26 52 56.56
0.75 2.70 1.41 - 5.17 0.90 33.22 52 70.00
1.00 2.30 1.19 - 4.46 0.78 33.92 52 82.25
1.25 1.95 1.04 - 3.68 0.64 32.51 63 96.75
1.50 1.93 1.00 - 3.71 0.65 33.55 63 108.87
1.75 1.72 0.88 - 3.37 0.60 34.67 63 122.12
2.00 2.09 1.13 - 3.86 0.66 31.57 84 150.56
2.25 1.80 1.04 - 3.12 0.51 28.16 95 174.95
2.50 1.49 0.87 - 2.55 0.41 27.72 95 211.80
Appendix III - 2
TAB LE III.2 Crax rubra griscomi cumulative density estimates for the areas away from different radial
distances from cenotes and aguadas (Chapter 4, Figure 4.1).
Away from
radial distance
from cenote
(km)
Cumulative
density estimate
(birds/km )
2
Confidence
Interval
(95%)
Standard
Error
% Coef.
Variation
Sample size
(transects
surveyed)
Sampling
effort
(km)
0.25 0.44 0.21 - 0.92 0.17 39.74 113 348.66
0.50 0.46 0.22 - 0.98 0.18 40.17 113 330.71
0.75 0.48 0.22 - 1.04 0.20 40.69 113 316.05
1.00 0.50 0.21 - 1.18 0.23 45.29 95 303.84
1.25 0.53 0.22 - 1.27 0.25 46.81 95 289.34
1.50 0.47 0.19 - 1.18 0.23 48.71 94 277.22
1.75 0.49 0.19 - 1.31 0.26 52.40 94 263.97
2.00 2.12 0.60 - 7.48 1.49 70.41 94 235.53
2.25 2.37 0.71 - 7.93 1.58 66.95 93 211.20
2.50 2.87 0.75 - 11.03 2.17 75.75 60 174.29
Appendix IV - 1
APPENDIX IV
HUMAN INFLUENCE ON THE COZUMEL CURASSOW DENSITY
TAB LE IV .1 Crax rubra griscomi cumulative density estimates for the areas within different linear distances
(along paths to the forest interior) from human settlements and human accesses to the forest interior (see Chapter
4, Figure 4.3).
Distance from
the source of
human
influence (km)
Cumulative
density estimate
(birds/km )
2
Confidence
Interval
(95%)
Standard
Error
% Coef.
Variation
Sample size
(transects
surveyed)
Sampling
effort (km)
0.5 0.00 20 10.00
1.0 0.00 21 20.22
1.5 0.00 21 30.72
2.0 0.30 0.04 - 2.16 0.39 130.45 86 72.82
2.5 0.18 0.03 - 1.11 0.21 114.94 92 121.22
3.0 0.26 0.06 - 1.01 0.20 78.18 92 169.87
3.5 0.70 0.29 - 1.68 0.32 46.06 92 216.75
4.0 0.90 0.37 - 2.21 0.43 47.58 92 261.15
4.5 0.87 0.43 - 1.79 0.33 37.36 92 289.05
5.0 0.90 0.48 - 1.69 0.30 32.82 110 302.65
5.5 0.85 0.45 - 1.62 0.29 33.82 110 321.65
6.0 1.01 0.57 - 1.79 0.30 29.70 112 331.80
6.5 0.98 0.55 - 1.75 0.30 30.18 112 341.80
7.0 0.96 0.53 - 1.72 0.29 30.67 112 351.80
7.5 0.91 0.50 - 1.65 0.28 31.45 113 371.36
8.0 0.89 0.48 - 1.63 0.28 31.74 113 377.86
8.5 0.88 0.47 - 1.62 0.28 32.02 113 383.86
8.795 0.87 0.47 - 1.61 0.28 32.12 113 386.09
Appendix IV - 2
TAB LE IV .2 Crax rubra griscomi density estimates for the areas at 1.5 km intervals from human settlements
and accesses to the forest interior (see Chapter 4, Figure 4.4).
Distance intervals
from the source of
human influence
(km)
Density
estimate
(birds/km )
2
Confidence
Interval
(95%)
Standard
Error
% Coef.
Variation
Sample size
(transects
surveyed)
Sampling
effort (km)
0.0 - 1.5 0.00 21 30.72
1.5 - 3.0 0.31 0.09 - 1.11 0.22 70.62 92 139.15
3.0 - 4.5 1.76 0.77 - 4.06 0.77 43.84 88 119.18
4.5 - 6.0 2.92 0.93 - 9.23 1.81 61.75 40 42.75
