An Approach to the Ecology of the Herpetofauna in Agroecosystems of the

Colombian Coffee Zone

Author(s): Julio Mario Hoyos-Hoyos, Paola Isaacs-Cubides, Nathalie Devia, Diana Maria Galindo-
Uribe and Andrés Rymel Acosta-Galvis
Source: South American Journal of Herpetology, 7(1):25-34. 2012.
Published By: Brazilian Society of Herpetology
DOI: http://dx.doi.org/10.2994/057.007.0103
URL: http://www.bioone.org/doi/full/10.2994/057.007.0103

BioOne (www.bioone.org) is a nonprofit, online aggregation of core research in the biological, ecological,
and environmental sciences. BioOne provides a sustainable online platform for over 170 journals and books
published by nonprofit societies, associations, museums, institutions, and presses.
Your use of this PDF, the BioOne Web site, and all posted and associated content indicates your acceptance of
BioOne’s Terms of Use, available at www.bioone.org/page/terms_of_use.
Usage of BioOne content is strictly limited to personal, educational, and non-commercial use. Commercial
inquiries or rights and permissions requests should be directed to the individual publisher as copyright holder.

BioOne sees sustainable scholarly publishing as an inherently collaborative enterprise connecting authors, nonprofit publishers, academic institutions,
research libraries, and research funders in the common goal of maximizing access to critical research.
South American Journal of Herpetology, 7(1), 2012, 25-34
© 2012 Brazilian Society of Herpetology

AN APPROACH TO THE ECOLOGY OF THE HERPETOFAUNA IN

AGROECOSYSTEMS OF THE COLOMBIAN COFFEE ZONE
JULIO MARIO HOYOS-HOYOS1,4, PAOLA ISAACS-CUBIDES1, NATHALIE DEVIA2,
DIANA MARIA GALINDO-URIBE3, AND ANDRÉS RYMEL ACOSTA-GALVIS1

1. Pontificia Universidad Javeriana, Departamento de Biología, Unidad de Ecología y Sistemática (UNESIS), A.A. 56710,
Bogotá, Colombia. E-mails: [email protected], [email protected], [email protected]
2. Universidad Nacional de Colombia, Instituto de Ciencias Naturales,
Ciudad Universitaria Calle 30 Carrera 45, Bogotá, Colombia. E-mail: [email protected]
3. Grupo de Ecofisiología, Comportamiento y Herpetología. Facultad de Ciencias, Departamento de Ciencias Biológicas.
Universidad de Los Andes, Bogotá, Colombia. E-mail: [email protected]
4. Corresponding author.

ABSTRACT. To establish the presence of amphibian and reptile species by cover and their diet in agroecosystems in the Colombian
coffee belt, we sampled four farms in the Departamento del Quindío and two in the Departamento del Valle del Cauca (Colombia)
from 2006 to 2008. Specimens were captured using both the VES technique and manual capture; each gastrointestinal sample was
analyzed to get the type of diet (carnivore, omnivore or herbivore) and the frequency of consumed items (detritus, animal parts,
plant material and minerals). Both the sensitivity index (IS) and the dominance index were calculated to find out whether each spe-
cies belongs to open areas or forest, and to determine niche breadth. Differences between stomach and intestine were analyzed by
using a Chi-square test. We captured 376 individuals of 19 species (nine frogs, three lizards, six snakes and one turtle) in crops, pas-
tures, forests and forest of Guadua angustifolia. Centrolenid frogs were found mainly in forests and Guadua, and Rhinella marina,
Leptodactylus colombiensis, Lithobates catesbeianus, two species of snakes and a turtle, in open areas. Gastrointestinal contents
showed that most of the species are carnivorous. This paper is the first report of the diet of these species in Colombia especially in
agroecosystems, being important to compare differences due to habitat transformation.

KEYWORDS. diet type, herpetofauna, coffee belt, agroecosystems.

INTRODUCTION on native species on the edge of patches, because both

tolerant and opportunistic species of the matrix colo-
Fragmentation of habitats and their consequent nize new areas (Murcia 2002). In addition, herbivory
loss caused by human actions, is considered by many increases causing a low survival rate, fecundity or
biologists as the greatest current threat to biodiver- growth of plant communities (Murcia 2002). In fact,
sity. This process reduces total habitat available and when the number of individuals decreases, ecological
landscape connectivity, affecting the habitat quality services (SE) such as pollination, dispersal and preda-
of species and isolating populations, making them tion will also decrease, causing a major fluctuation of
more vulnerable to extinction (Murcia 1995). individuals that may produce decline and even extinc-
The Colombian coffee belt is a region of trans- tion of species populations (Murcia 2002). Changes
formed and fragmented landscapes owing to the in biotic and abiotic factors may have an effect on
diversity of productive systems, where the forest is predation activities of species making the presence of
surrounded by an agricultural matrix, or restricted opportunistic and tolerant species easier and driving
to isolated remnants where extensive crops occur sensitive species to extinction, or may even promote
(CARDER 2004). a reaction chain inside the communities that could
Agriculture in the coffee belt produces deleterious alter the ecosystem balance (Murcia 2002). Habitat
effects owing to area fragmentation, and increases destruction could alter most species but not influence
the vulnerability associated with genetic uniformity equally all species (Pineda et al. 2008), and their vul-
of monoculture. nerability depends on reproductive mode, foraging
Therefore, knowledge about how organisms inter- strategy, body size (Duellman and Trueb 1994) and
act in agricultural ecosystems is important for both diet breadth (Suazo-Ortuño et al. 2008).
integrated management programs and sustainable ag- One of the important aspects that could be af-
riculture (Peltzer et al. 2004). fected by habitat loss is feeding habits; this includes
Some important effects of fragmentation are per- behavioral, physiological and morphological charac-
ceived at the level of community ecological process- ters that are important in population dynamics of spe-
es: it increases competition, parasitism and predation cies and that will determine the interactions between
26 Ecology of the Herpetofauna in Colombian Coffee Zone

competitors and other upper trophic levels (Hirai and bimodal precipitation regime with values lower than
Matsui 2002, Lima and Magnusson 2002). Informa- 1800mm per year, and with humid periods extending
tion on the diet of amphibians and reptiles is impor- from March to May, and from September to Novem-
tant to understand how herpetofaunal interactions af- ber; the dry season occurs from January to February
fect the community (Hirai and Matsui 2002). Study and from June to August. The average temperature is
on the diet of Colombia herpetofauna are scarce (e.g., between 18°C and 21°C (IGAC 1996). The rainiest
Daza-Vaca and Castro-Herrera 1999, Botero-Trujillo period occurs from April to May and from October to
2006, Muñoz-Guerrero et al. 2009), even more in November (IGAC 1996).
productive systems. For that reason, the aims of this
paper are to establish the presence of amphibian and Methodology: We used the Visual Encounter Survey
reptile species in a forest of Guadua angustifolia, a method (VES, Crump and Scott 1994) to sample all
riparian remnant forest, crops, grasses and human set- the different covers present in the farms (G. angusti-
tlings, and to study their diet as possibly representing folia, forest, crops, grasses and around the houses).
a more general situation in the agroecosystems of the Given the logistic difficulties, the surveys were not
Colombian coffee belt. always carried out by the same number of people in
the same months of each year; for that reason, we
have an average of 210 hours/people, taking into ac-
MATERIALS AND METHODS count that every survey was carried out by at least
two people.
Area of Study: The present study was carried out The observed individuals were collected by manu-
in the Colombian coffee belt in the departamentos al capture method, deposited in a plastic bag and sac-
(Provinces) of Quindío and Valle del Cauca, which rificed following the ethic protocols using a saturated
are characterized by a history of interventions to the solution of chloretone and ethanol in water (Angulo
original native vegetation by the introduction of live- et al. 2006). We did not use stomach flushing given
stock and agriculture, which has reduced the natural the small size of some frogs. They were fixed inject-
cover to areas of Guadua angustifolia to preserve the ing formalin 10% in several points of the ventral re-
water resources and some patches of remnant vegeta- gion (Heyer et al. 1994, Acosta 2002a) to avoid prey
tion, which still maintain the biodiversity of the re- decomposition; then, they were maintained in a humid
gion (IGAC 1996). camera dampened with 10% formalin for eight hours.
Four farms were visited in Quindío: La Ramada Finally, the samples were preserved in 70% ethanol,
(in June of 2006 and April of 2008) (04°34’97.7”N, and were deposited in the Museo de Historia Natural
75°49’89.6”W; 1200 m), La Floresta (in November de la Pontificia Universidad Javeriana (MUJ).
of 2006, in June of 2006 and of 2007) (04°35’70.6”N, We made a lateral cut in the left flank of the pre-
75°50’13.4”; 1180 m), Tesalia Baja (in September served specimens, from the armpit to the groin; then,
of 2006 and in November of 2007) (04°35’0.85”N, both the stomach and the whole intestine were cut
75°40’0.12”W; 1620 m) and Lusitania (in June of laterally, including the cloaca, and the contents was
2007) (04°41’50.2”N, 75°50’0.52”W; 1200 m), preserved in 70% ethanol, preferring this method be-
and we visited two farms in the Valle del Cauca: La cause major data can be obtained (Parmelee 1999,
Comarca (in September of 2006) (04°41’45.4”N, Guayasamin et al. 2004).
75°46’19.9”W; 1230 m) and El Topacio (in April of For both the gastrointestinal content and feces
2008) (04°40’24”N, 75°43’60”W; 1560 m). Each analysis, we used the occurrence method proposed
farm consists of a mosaic of productive systems based in ichthyological studies by Hyslop (1980), which
mainly on coffee, banana, and pastures for livestock, registers, in frequency terms, the number of stom-
with riparian forest of Guadua angustifolia, second- achs and intestines that contain individuals or items
ary forest and some remnants of original forest that of a determined category (in our case invertebrates,
still maintain some of the ecological characteristics detritus, plant material and minerals). This method
for the fauna and the flora. is commonly used because it allows to obtain, in an
The Colombian coffee belt region is influenced by easy and simple way, a quick qualitative vision of the
the intertropical convergence zone, by the mountain individual nutritional spectrum for categorical data
relief (influenced by both the humid air of the cordil- (Hyslop 1980). The preys were determined to the
lera Occidental and the Cauca river basin) and by the smallest possible taxonomic category and the other
atmospheric circulation (IGAC 1996). It presents a types of material were classified as detritus, plant
 Hoyos-Hoyos, J. M. et al. 27

material and/or mineral. In the same way, to deter-

mine the type of diet (carnivorous, omnivorous or
herbivorous), the content of the sample was placed
in a Petri dish with a millimetric sheet adhered to the Where: Pi is the proportional use of the resource i
bottom to calculate the percentage occupied by the (orders) and n is the total number of categories (to-
category. The contents were grouped, and the number tal orders). The value varies from 1 (exclusive use of
of free squares were counted, in order to calculate the single prey type), to n (all prey types used equally,
percentage occupied. This article is part of a mega Pianka 1973).
project framed by an Excellence Center of Research
“Centro de Investigaciones y Estudios en Biodiver-
sidad y Recursos Genéticos (CIEBREG)”. We were RESULTS
given permission by the Ministry of Environment to
collect and sacrifice specimens, and it was approved We captured a total of 376 individuals belonging
by the Bioethics Committee of the Pontificia Univer- to 19 species in the area: nine species of frogs, three
sidad Javeriana. Both collection and sacrifice permis- of lizards, six of snakes and one of turtle. Snakes, the
sion and approval was general and not individual. turtle and the lizard Basiliscus galeritus, were not in-
cluded in the analysis of diet, because we only cap-
Data Analyses: To establish if the species found be- tured one specimen of each species (Table 1).
longed to open or forest areas, we used the Sensibility According to the different vegetation covers
Index (IS) proposed by Pineda and coworkers (2008) present in the area, 66% of amphibians and reptiles
and calculated according to the following equation: were caught in open areas (crops, grasses, around
the houses) and 34% in forest areas (G. angustifolia,
G. angustifolia – forest, forest). The percentage of
the specimens captured in open areas was as follows:
52% in grassland, 13% surrounding houses, and 1%
where: H1 is the number of captured individuals in in crops (Fig. 1), and in forests: 16% were captured
habitat one (open areas) and H2 the number of cap- in G. angustifolia, 10% in forest and 8% in forest–
tured individuals in the comparison habitat (forest). G. angustifolia. In terms of richness, covers that pre-
The index goes from values of -1 to 1, being -1 an sented the greatest number of species were grasses
indicator of low sensibility, 0 neutrality, and 1 high (13 species) and G. angustifolia with eight species
sensibility (Pineda et al. 2008). In this case, the open (Fig. 1).
areas include crops, grasses and the surroundings Taking into account the IS values, centrolenids,
of the houses and the forest areas of G. angustifolia B. galeritus, Chironius monticola, Imantodes cen-
(G. angustifolia – forest, forest). choa, Leptodeira annulata and Sibon nebulatu
According to the diet found in gastrointestinal showed values of 1; on the other hand, R. marina,
contents, the presence of the different items was ex- L. colombiensis, L. catesbeianus, the snakes Eryth-
pressed in frequency of occurrence by category (Hys- rolamprus bizona, Leptophis depressirostris and the
lop 1980). We used a Chi-square test to show differ- turtle Chelydra serpentina showed an IS of -1. D. co-
ences between stomach and intestine (Zar 1999). The lumbianus, C. fraterdanieli and A. auratus had values
diet type was classified using the percentage of data, close to -1; A. antonii, and the species of the Pris-
following Cooper and Vitt (2002) in which a species timantis genus, were found in both open and forest
is considered omnivorous if at least 10% of its diet areas (Table 1).
is plant material, and carnivorous or herbivorous, if With regard to the analysis of the diet, the Chi-
it contains 90% or more of animals or plant material square test showed that there are no significant differ-
respectively. This classification could be considered ences among species between stomach and intestine
as arbitrary, but it allows one to exclude the acciden- contents (Table 2).
tal consumption of plant material for the case of the The most abundant prey categories were Cole-
carnivores (Cooper and Vitt 2002). To estimate the opterans, Hymenoptera, Hemiptera, Arachnidae and
trophic niche breadth and to determine if the diet type Diptera, being Coleoptera the one with the high-
is generalist or specialist (Pianka 1973), the recipro- est values of abundance. We also found 44 contents
cal dominance of Simpson index was calculated (B), with nematodes, 115 with minerals and 129 with
following the equation: plant material (Fig. 2). The species were categorized
28 Ecology of the Herpetofauna in Colombian Coffee Zone

TABLE 1. Species of frogs and reptiles found in agroecosystems of the Colombian coffee belt.

Family Species n Hábitat* IS index

Hylidae Dendropsophus columbianus 146 A, OA -0,71
Bufonidae Rhinella marina 48 OA -1,00
Strabomantidae Pristimantis achatinus 81 F, OA 0,31
Pristimantis palmeri 12 F 0,83
Dendrobatidae Colostethus fraterdanieli 14 L -0,71
Centrolenidae Centrolene prosoblepon 15 F 1,00
Hyalinobatrachium colymbiphyllum 12 F 1,00
Leptodactylidae Leptodactylus colombiensis 2 OA -1,00
Ranidae Lithobates catesbeianus 1 OA -1,00
Polychrotidae Anolis antonii 30 OA, F 0,53
Anolis auratus 6 OA, F -0,67
Corytophanidae Basiliscus galeritus 1 F 1,00
Colubridae Chironius monticola 1 F 1,00
Erythrolamprus bizona 1 OA, F -1,00
Imantodes cenchoa 1 F 1,00
Leptodeira annulata 1 F, OA 1,00
Leptophis depressirostris 1 F -1,00
Sibon nebulatus 1 OA, F 1,00
Chelydridae Chelydra serpentina 1 A -1,00

* (A: aquatic; OA: open areas; F: forest; L: litter; H: houses).

as carnivores, although D. columbianus presented a from seven to 17 Orders for the most abundant ones.
high frequency of plant material as well as minerals. The anoles and D. columbianus showed high values
We took the three most frequent categories and val- related to the number of Orders present in their diet
ues of the S index to determine the dietary preference (Table 3), and D. columbianus was the species with
of the species (see prey consumed and S in Table 2): the highest number of Orders consumed (n = 17).
C. fraterdanielli and P. achatinus presented low val- Colostethus fraterdanielli was found in the litter
ues, although the spectrum of prey consumed varied where it is very difficult to see, we only observed

FIGURE 1. Abundance and richness of species per cover sampled (Guadual = Guadua angustifolia).
 Hoyos-Hoyos, J. M. et al. 29

TABLE 2. Diet of the anurans and reptiles obtained from the gastrointestinal contents.

Frequency
Percentage Prop. Prey consumed
Stomach Intestine Feces
Species n Chi/D.f.
No
Freq % Freq. % Freq. % % pre. % det. % p.m. Type S 1 2 3
orders
Dc 146 111 32 215 63 16 5 11.2/13 66.3 32.4 1.24 C 5.99 14 CO HE AR
Pa 81 102 41 136 55 9 4 21.5/13 83 15.7 1.25 C 4.97 14 CO HY HE
Pp 12 21 58 12 33 3 8,3 4.27/6 80 20 0 C 5.55 7 HY HE AR
Cf 14 37 52 33 46 1 1,4 3.64/6 100 0 0 C 2.9 7 HY CO DI
Cp 15 8 31 14 54 4 15 — 70 30 0 C — 3 AR DI AC
Hc 11 6 30 12 60 2 10 — 50 50 0 C — 4 HY CO DI
Lc 2 7 77 2 23 — — — 100 0 0 C — 3 HE CO AR
Lca 1 — — — — — — — 100 0 0 C — 3 HY OR HE
Aa 31 31 26 39 33 49 41 5.17/10 100 0 0 C 6.46 11 CO AR HY
Aau 6 3 20 5 33 7 46 3.75/5 100 0 0 C 5.33 6 HY AR —

* (Dc: Dendropsophus columbianus, Pa: Pristimantis achatinus, Pp: Pristimantis palmeri, Cf: Colostethus fraterdanielli, Cp: Centrolene
prosoblepon, Hc: Hyalinobatrachium fleischmanni, Lc: Leptodactylus colombiensis, Lca: Lithobates catesbeianus, Aa: Anolis antonii,
Aau: Anolis auratus. Pre: prey, Det: detritus, P.M.: plant material, Type O: omnivorus, C: carnivorus, Prop.: proportion, S: Simpson index.
Prey consumed, CO: Coleoptera, HY: Hymenoptera, HE: Hemiptera, AR: Araneae, DI: Diptera, AC: Acari).

FIGURE 2. Frequencies of categories found in both the gastrointestinal tracts and feces as a whole.
30 Ecology of the Herpetofauna in Colombian Coffee Zone

TABLE 3. List of frequency of categories found by species.

Species Aa Aau Dc Pa Pp Cf Cp Hf Lc Lca TOTAL

Categories 12 5 15 13 7 7 4 5 5 4
Coleoptera 24 1 43 35 2 7 1 2 3 0 193
Hymenoptera 13 2 27 35 5 25 0 2 0 3 187
Hemiptera 12 0 40 14 4 3 0 1 5 1 97
Araneae 15 2 38 14 3 3 3 0 1 0 88
Diptera 6 1 18 6 0 5 1 2 0 0 42
Lepidoptera 10 0 9 7 0 0 0 0 0 0 33
Orthoptera 4 1 7 2 3 0 0 0 0 1 21
Acari 2 0 2 1 1 0 1 1 0 0 13
Chilopoda 0 0 0 1 0 1 0 0 0 0 4
Dermaptera 3 0 1 0 0 0 0 0 0 0 5
Odonata 0 1 2 0 0 0 0 0 0 0 3
Diplopoda 0 0 0 1 0 0 0 0 0 0 2
Gastropoda 0 0 0 0 0 2 0 0 0 0 3
Collembola 1 0 1 0 0 0 0 0 0 0 2
Dictioptera 0 0 0 2 0 0 0 0 0 0 4
Crustacea 1 0 1 0 0 0 0 0 0 0 2
Collembola 1 0 0 1 1 0 0 0 0 0 3
Psocoptera 0 0 1 0 0 0 0 0 0 0 2
Blattodea 0 0 1 0 0 0 0 0 0 0 2
Myriapoda 0 0 0 0 0 0 0 0 0 0 2
Isoptera 0 0 0 1 0 0 0 0 0 0 1
Pseudoescorpion 0 0 0 1 0 0 0 0 0 0 1
Nematoda 3 1 11 9 1 3 1 0 1 0 44
Plant material 4 2 42 30 2 5 4 4 0 2 129
Minerals 3 0 32 38 5 7 1 0 1 0 115
Detritus 18 1 66 49 9 10 7 6 0 0 201

* (Aa: Anolis antonii, Aau: Anolis auratus, Dc: Dendropsophus columbianus, Pa: Pristimantis achatinus, Pp: Pistimantis palmeri, Cf:
Colostethus fraterdanielli, Cp: Centrolene prosoblepon, Hc: Hyalinobatrachium fleischmanni, Lc: Leptodactylus colombiensis, Lca:
Lithobates catesbeianus).

them in the pasture once, for that reason the sensi- DISCUSSION
tivity index was skewed toward open areas, although
they were heard vocalizing in forests and G. angusti- According to our findings, most of the herpeto-
folia. On the other hand, Anolis antonii preferred for- fauna were found in open areas, being the grassland
est areas and A. auratus open areas. the most representative (52% of the samples) in terms
It is important to remark that Lithobates cates- of abundance and richness.
beianus, of which we only captured one specimen, Taking into account the IS values, centrolenids,
had in its gastrointestinal content fragments of He- B. galeritus, Chironius monticola, Imantodes cen-
miptera, Orthoptera and Hymenoptera; in this last choa, Leptodeira annulata and Sibon nebulatu were
case we found a queen ant of the species Ectatoma exclusive to the forest. Rhinella marina, L. colom-
tuberculatum. biensis, L. catesbeianus, the snakes Erythrolam-
The gastrointestinal contents by cover did not vary prus bizona, Leptophis depressirostris and the turtle
and maintained the trend of prey consumption of the Chelydra serpentina were exclusive to open areas.
Orders Coleoptera, Hymenoptera, Hemiptera and A. antonii, and the species of the Pristimantis ge-
Araneae (Fig. 3), finding for both grassland and areas nus, could be found in both open and forest areas
around the houses higher values for D. columbianus (Table 1).
than for the other species, given that these species are Our results support other studies (Acosta, 2002a)
abundant and live nearly exclusively in these areas. showing that D. columbianus is a species widely tol-
Instead, data are very limited in both crops and forest- erant to intervention, living in degraded areas, and
guaduales due to the low number of individuals found that might eventually be found in edges of forests, but
in these covers. is less abundant than in water bodies of open areas.
 Hoyos-Hoyos, J. M. et al. 31

FIGURE 3. Prey frequencies per species found in the gastrointestinal tract. We only show those for the most common Orders.

The two species of centrolenids and P. palmeri, are We would rather use the frequency (defined as
species of forests. P. palmeri is a species that has been a measure of the number of times that a particular
reported in secondary forests, seldom surpassing the prey or item occurs in predatory species; Hyslop
edges of forests (Acosta 2002b). Given that centrolenids 1980, Parmelee 1999, Zar 1999) to describe species
require oxygenated water to lay their eggs, they need diet, because it is more conservative in the infor-
forest covers where conditions are more stable and the mation that it shows, considering that the digestive
water is less polluted. P. achatinus prefers forests, but it processes fragment individuals in parts at different
could also be found in open areas, being the only one in rates depending on the type of prey, hindering the
crops, confirming the generalistic nature of this species. calculation of the abundance and real volume (Zug
This paper is the first one to report the type of diet 1993).
of species in agroecosystems on the Colombian cof- P. achatinus presented a tendency to have a pref-
fee belt, showing that most of them feed on arthro- erential diet, as demonstrated by the low values ob-
pods, consuming a wide range of prey. There have not tained by the dominance index (see Table 2) with
been studies of frogs and reptiles diet of D. colum- respect to the number of orders obtained from their
bianus, P. achatinus, C. fraterdanielli and A. antonii gastrointestinal tracts; however, the spectrum of prey
despite that these are the most abundant and common consumed by all species is wide showing preference
species in this region. for coleopterans and hymenopterans.
32 Ecology of the Herpetofauna in Colombian Coffee Zone

Muñoz et al. (2007) reported the orders Orthop- catesbeianus (bull frog), being important because
tera, Arachnida, Coleoptera, Hymenoptera and Dip- L. catesbeianus is an introduced species, using the
tera as the most common and frequent in the stomach prey in a different way than the native species. Simi-
contents of Dendropsophus microcephalus, Scarthyla larly, it is important to mention the presence of plant
vigilans, Hypsiboas pugnax and Scinax rostratus and, material in gastrointestinal contents, because it can be
to a lesser extent, Blatteridae, Isoptera, Hemiptera considered accidental (Zug and Zug 1979, Zug 1993),
and Lepidoptera. While we did not find isopterans, and maybe the presence of plant material helps elimi-
the orthopterans were the most abundant and, addi- nate intestinal parasites, providing fibers to help to
tionally, we extracted mites, crustaceans, psocopter- crush hard parts as those of coleopterans, and contrib-
ans and Collembola. We found families of hemipter- utes as an additional water source during periods of
ans, spiders and flies in D. columbianus that were not drought (Anderson et al. 1999).
found in other anurans, nevertheless the greater pro- For those items identified to genus and species
portion of prey found corresponds to himenopterans (beetles and hymenopterans), we observed differ-
and spiders like those reported in other studies for the ences between the composition of prey consumed
family Hylidae (Daza-Vaca and Castro-Herrera 1999, in the forest and in open areas, apparently because
Parmelee 1999, Menéndez 2001). This shows that the the offer is different. This makes it difficult to know
greatest quantity of prey of a certain type will have an whether the herpetofauna has a preference for eating
aquatic habit (Parmelee 1999), which is in agreement specimens of a specific family or genus of arthropods.
with the habitat preferred by this species. However, the work with gastrointestinal contents
We may say that P. achatinus has preferences could be biased given that the size and biomass of
related to habitat types (forests), because it con- beetles and hymenopterans may exceed that of other
sumes few flies but a lot of ants of different families, animals (Parmelee 1999, Menéndez 2001).
while we were able to recognize that P. palmeri and Our findings about the two species of Anolis (an-
C. fraterdanielli prefer to live closest to the ground tonii and auratus) are the first ones reported in the lit-
level because we found myriapods in their gastroin- erature, given that only the diets of Anolis oculatus in
testinal system. Dominica, and Anolis capito in Nicaragua have been
The largest proportion of prey eaten by Pristi- reported (Vitt and Zani 2005). Our species presented
mantis (as a genus) corresponds to beetles and hy- a high number of items of spiders, lepidopterans, bee-
menopterans, as reported in other studies (Parmelee tles and Hymenoptera, while A. oculatus consumes
1999 and references therein, Lima and Magnusson crickets and A. capito prefers crickets, homopterans,
2002, Guayasamin et al. 2004), but in species of larvae, cockroaches, beetles and Diplopoda, eating a
this genus in the Amazonian Cuzco, most of their low number of prey, mostly of large size.
diet consists of beetles and orthopterans (Parmelee We found species that prefer open areas or forests
1999), contrasting with our results in which we did in agroecosystems with a diet type mostly carnivo-
not find orthopterans but hemipterans (ranked fourth rous biased towards coleopterans, himenopterans and
in frequency). hemipterans. We can conclude also that the diversity
Comparing dendrobatid (sensu lato) species, we of amphibians and reptiles in the coffee belt is very
see that C. fraterdanielli is specialized in hymenopter- low, although certain species remain due to the inter-
ans (Table 3), as has been shown for C. marchesianus vention of the agroecosystem. The species are highly
in Ecuador. Although its diet consists of collembo- specialized in this agroecosystem, preferring cole-
lans, mites and beetles too, the same species in the opterans, hemipterans, hymenopterans and Araneae;
Peruvian Amazon consumes ants and mites (Toft however, thess preferences are probably due to the
1980). higher availabilty of these taxa, although more stud-
Since our report is the first known about the diet of ies are necessary to reach this conclusion (Duellman
the centrolenid species C. prosoblepon (spiders and and Trueb 1994).
beetles) and H. colymbiphylum (beetles and flies), we Finally, additional work should be done taking
were not able to compare it with other studies about into account both weight and volume of gastrointesti-
this family, but we can notice that their feeding habits nal contents, to determine more accurately which are
are not very different from those of the other species the most abundant items (number and biomass), and
described above. diet preference, and identifying arthropods in agro-
It is worth noting the presence of a queen ant of ecosystems and their relationship with both frog and
the species Ectatoma tuberculatum in Lithobates lizard preferences.
 Hoyos-Hoyos, J. M. et al. 33

RESUMEN Corporación autónoma regional de Caldas. Subdirección

Administración de los Recursos Naturales. Grupo de
Ordenamiento Territorial. Colombia.
Dieta de la herpetofauna presente en el eje cafete-
Anderson, A. M., D. A. Haukos, and J. T. Anderson. 1999. Diet
ro colombiano. Se determinó presencia por cobertura Composition of Three Anurans from the Playa Wetlands of
y el tipo de dieta de la herpetofauna en los agroecosis- Northwest Texas. Copeia, 2:515-520.
temas en el eje cafetero colombiano. Se visitaron cua- Botero-Trujillo, R. 2006. Anuran predators of scorpions: Bufo
tro fincas, dos en el Quindío y dos en el Valle del Cau- marinus (Linnaeus, 1758) (Anura: Bufonidae), first known
natural enemy of Tityus nematochirus Mello-Leitão, 1940
ca, del año 2006 al 2008. Se capturó manualmente (Scorpiones: Buthidae). Revista Ibérica de Aracnología,
la herpetofauna observada por medio de las técnicas 13:199-202.
VES; se analizó el contenido gastrointestinal para de- Angulo, A., J. V. Rueda-Almonacid, J. V. Rodríguez-Mahecha,
terminar el tipo de dieta (carnívoro, omnívoro o her- and E. La Marca (Eds). 2006. Técnicas de inventario y
monitoreo para los anfibios de la región tropical andina.
bívoro) y la frecuencia de ítems consumidos (detritos, Conservación Internacional. Serie Manuales de Campo Nº 2.
material vegetal, partes de animales y minerales). Se Panamericana Formas e Impresos S.A., Bogotá D.C.
calcularon el índice de sensibilidad (IS), para deter- Cooper, W. E. and L. J. Vitt. 2002. Distribution, extent, and
minar si cada especie es de zonas abiertas o de bosque evolution of plant consumption by lizards. Journal of the
Zoological Society of London, 257:487-517.
y el índice de dominancia para determinar amplitud Corporación Autónoma Regional de Risaralda (CARDER). 2004.
de nicho. Se analizaron diferencias entre estómago Ecorregión Eje Cafetero: un territorio de oportunidades.
e intestino por medio de una prueba Chi-cuadrado. Segunda Edición.
Se capturaron 376 individuos de 19 especies (nueve Crump, M. L. and N. J. Scott. 1994. Visual Encounter Surveys.
Measuring and Monitoring Biological Diversity. Standard
de ranas, tres de lagartos, seis de serpientes y una de
Methods for Amphibians. In: W. Heyer et al. (Eds.).
tortuga), en cultivos, pastos, bosques y “guaduales”. Smithsonian Institution Press, Washington D.C., U.S.A.
Los centrolénidos son exclusivos de bosque, y de zo- Da Silva, H. R. and M. C. de Britto-Pereira. 2006. How much
nas abiertas Leptodactylus colombiensis, Lithobates fruit do fruit-eating frogs eat? An investigation on the diet of
Xenophyla truncata (Lissamphibia: Anura: Hylidae). Journal
catesbeianus, dos especies de serpientes y la tortuga.
of Zoology, 270:692-698.
Se determinó que la carnivoría es el tipo más común Daza-Vaca, J. D. and F. Castro-Herrera. 1999. Hábitos alimenticios
de dieta, con un consumo de coleópteros e himenóp- de la rana toro (Rana catesbeiana) Anura: Ranidae, en el Valle
teros. Este es el primer reporte sobre la dieta de estas del Cauca, Colombia. Revista de la Academia Colombiana de
especies en el país especialmente en agroecosistemas, Ciencias, 23:265-274.
Duellman, W. E. and L. Trueb. 1994. Biology of Amphibia. The
siendo un reporte importante para comparar diferen- John Hopkins University Press. Baltimore, USA. 670p.
cias debidas a la transformación del hábitat. Guayasamin, J. M., E. Bonaccorso, P. A. Menéndez, and
M. Bustamante. 2004. Morphological variation, diet,
and vocalization of Eleutherodactylus eugeniae (Anura:
ACKNOWLEDGMENTS Leptodactylidae) with notes on its reproduction and ecology.
Herpetological Review, 35:17-23.
We want to thank everybody in the farms that gave us fa- Hero, J. M. and M. Stoneham. AmphibiaWeb: Information on
cilities to work there; to Sergio Moreno for helping us in both amphibian biology and conservation. http://amphibiaweb.
the gastrointestinal analysis, and in the field; to Lina Parada, Fer- org/cgi/amphib_query?where-genus=Bufo&where-
nanda Cantillo, Cristina Cepeda and Carolina Piedrahita for their species=marinus. Consulted Jun 19 (2008).
help in the field. Likewise, to J. Parmelee, L. Vitt, P. Taylor and Heyer, R. W., M. A. Donelly, R. W. McDiarmid, L. C. Hayek,
J. Caldwell for their personal comments. This paper is part of the and M. S. Foster. (Eds.). 1994. Measuring and Monitoring
project “Valoración de bienes y servicios de la biodiversidad en el biological diversity: standard methods for amphibians.
eje cafetero: Complejo ecorregional Andes del Norte de Colom- Smithsonian Institution Press, Washington, D.C.
bia” carried out by the Excellence Center of Research “Centro de Hirai, T. and M. Matsui. 2002. Feeding Ecology of Bufo japonicus
Investigaciones y Estudios en Biodiversidad y Recursos Genéti- fortnosus from the Montane Region of Kyoto, Japan. Journal
cos (CIEBREG)” with the financial support of Colciencias and of Herpetology, 36:719-723.
the Pontificia Universidad Javeriana of Bogotá. We want to thank Hyslop, E. J. 1980. Stomach contents analysis – a review of
the Ministerio del Medio Ambiente, Vivienda y Desarrollo Ter- methods and their application. Journal of Fish Biology,
ritorial of Colombia for giving us collection authorization. 17:411-429.
Instituto Geográfico Agustín Codazzi (IGAC). Quindío y
Risaralda: características geográficas. Subdirección de
Geografía. Instituto Geográfico Agustín Codazzi.
REFERENCES Krakaguer, T. 1968. The Ecology of the Neotropical Toad, Bufo
marinus, in South Florida. Herpetologica, 24:214-221.
Acosta, A. R. 2002a. Ranas, Salamandras y Caecilias (Tetrapoda: Lima, A. and W. E. Magnusson. 2002. Does foraging activity
Amphibia) de Colombia. Biota Colombiana, 1:289-319. change with ontogeny? An assessment for six sympatric
Acosta, A. R. 2002b. Caracterización herpetofaunística de las species of postmetamorphic litter anurans in central
cuencas de los ríos Tapias y Tareas (Departamento de Caldas). Amazonia. Journal of Herpetology, 34:192-200.
34 Ecology of the Herpetofauna in Colombian Coffee Zone

Menéndez, P. A. 2001. Ecología trófica de la comunidad de anuros Coffee Agroecosystems of Veracruz, Mexico. Conservation
del Parque Nacional Yasuní en la Amazonía Ecuatoriana. Biology, 19:400-410.
Trabajo de Grado de Maestría – Facultad de Ciencias Exactas Suazo-Ortuño, L., J. Alvarado-Díaz, and M. Martínez-Ramos.
y Naturales, Pontificia Universidad Católica del Ecuador. 2008. Effects of Conversion of Dry Tropical Forest to
Muñoz-Guerrero, J., V. H. Serrano, and M. P. Ramirez-Pinilla. agricultural Mosaic on Herpetofaunal Assemblages.
2007. Uso de microhábitat, dieta y tiempo de actividad en Conservation Biology, 22:362-374.
cuatro especies simpátricas de ranas hílidas neotropicales Toft, C. A. 1980. Feeding ecology of thirteen syntopic species
(Anura: Hylidae). Caldasia, 29:413-425. of anurans in a seasonal tropical environment. Oecologia,
Murcia, C. 1995. Edge effects in fragmented forest implications 45:131-141.
for conservation. Trends in Ecology and Evolution, 10:58-62. Toft, C. A. 1981. Feeding ecology of Panamanian litter anurans:
Murcia, C. 2002. Ecología de la polinización. 493p. In: Ecología patterns in diet and foraging mode. Journal of Herpetology,
y conservación de bosques neotropicales. M. R. Guariguata 15:139-144.
and G. H. Kattan (Eds.). Universidad Nacional Autónoma, Vitt, L. J. and P. A. Zani. 2005. Ecology and Reproduction of
Cartago, Costa Rica. Anolis capito in Rain Forest of Southeastern Nicaragua.
Parmelee, J. R. 1999.Trophic ecology of a tropical anuran Journal of Herpetology, 39:36-42.
assemblage. Scientific Papers Natural History Museum The Wells, K. D. 2007. The ecology and behavior of amphibians.
University of Kansas, 11:1-59. University of Chicago Press. USA.
Peltzer, P., G. Bock, R. Tardito, and R. Lajmanovich. 2004. Zar, J. H. 1999. Biostatistical analysis. Cuarta edición. Prentice-
Effects of habitat loss and fragmentation on anuran in Espinal Hall Inc., New Jersey, USA.
Eco-region of Argentina: a GIS approach. Froglog, 63:3-4. Zug, G. R. and P. B. Zug. 1979. The marine toad, Bufo marinus:
Pianka, E. R. 1973. The structure of lizard communities. Annual a natural history resumé of native populations. Smithsonian
Review of Ecology and Systematics, 4:53-74. Contributions to Zoology, 284:1-58.
Pineda, E., C. Moreno, F. Escobar, and G. Halfter. 2008. Frog, Zug, G. R. 1993. Herpetology: an introductory biology of
Bat, and Dung Beetle Diversity in the Cloud Forest and amphibian and reptiles. Academic Press. London, UK.

Submitted: 18 February 2011

Accepted: 23 December 2011