SciELO - Scientific Electronic Library Online

vol.6 issue1Cartographic basis for storage and spatial analysis of data on diversity of palm trees in a stretch of Mata Atlântica, Ubatuba-SPPlant inhabiting mites (Acari) of the "Estação Ecológica de Paulo de Faria", State of São Paulo, Brazil author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand



  • English (pdf)
  • Article in xml format
  • How to cite this article
  • SciELO Analytics
  • Curriculum ScienTI
  • Automatic translation


Related links


Biota Neotropica

On-line version ISSN 1676-0611

Biota Neotrop. vol.6 no.1 Campinas  2006 



Detectability of capybaras in forested habitats


Detectabilidade de capivaras em habitats florestais



Gustavo Romeiro Mainardes PintoI; Kátia Maria Paschoaletto Micchi de Barros FerrazII,IV; Hilton Thadeu Zarate do CoutoII; Luciano Martins VerdadeIII

IInstituto Brasileiro do Meio Ambiente e dos Recursos Naturais Renováveis, Gerência Executiva de Santa Catarina, Av. Mauro Ramos 1113, 88020-3003 Florianópolis, Santa Catarina, Brasil
IILaboratório de Métodos Quantitativos, Departamento de Ciências Florestais, ESALQ, Universidade de São Paulo, Cx. P. 09, 13418-900 Piracicaba, São Paulo, Brasil
IIILaboratório de Ecologia Animal, Departamento de Ciências Biológicas, ESALQ, Universidade de São Paulo, Cx. P. 09, 13418-900 Piracicaba, São Paulo, Brasil
IVAutor para correspondência:




Direct count has been commonly used as an abundance index to estimate wildlife population size. However, systematic errors in sample-based estimators are common in sampling animal populations. At this study we aimed to estimate capybara's observability, through a detectability index in forested habitats. Sampling surveys of capybaras population was obtained by direct count (abundance index) and also by complete count (census). The average detectability index of capybaras in forested habitats was 0.63 ± 0.32 for a single observer. The variability in the detectability index among habitats was due to the presence of more or less dense vegetation. This information might be useful for management purposes of the species in the agroecosystems of southeastern Brazil where species is widespread and overabundant.

Key words: Hydrochoerus hydrochaeris, capybara, abundance index, direct count, detectability index


O método de contagem direta de animais tem sido freqüentemente utilizado como índice de abundância para estimar o tamanho de populações silvestres. Entretanto, erros sistemáticos de parâmetros baseados em amostras são comuns em amostragens de populações de animais silvestres. Neste estudo, nós estimamos a observabilidade de capivaras, através de um índice de detectabilidade em habitats florestais. As amostragens da população foram realizadas através da contagem direta (índice de abundância) e da contagem total (censo). O índice médio de detectabilidade de capivaras do observador em habitats florestais foi de 0,63 ± 0,32. A variabilidade no índice de detectabilidade entre os habitats ocorreu devido à vegetação ser mais ou menos densa. Esses resultados deverão ser úteis para o manejo da espécie em agroecossistemas do sudeste brasileiro onde a espécie é amplamente distribuída e abundante.

Palavras-chave: Hydrochoerus hydrochaeris, capivara, contagem direta, índice de abundância, índice de detectabilidade




Direct count has been commonly used as an abundance index to estimate wildlife population size (Lancia et al. 1996, Thompson et al. 1998, Williams et al. 2002). Although, systematic errors in sample-based estimators are common in sampling animal populations, good estimates of population size are essential for wildlife management and conservation, as well as control (Williams et al. 2001). Unfortunately, as Williams et al. (2002) have pointed out, the issue of variable detection rate has received insufficient emphasis in both observational and experimental studies in ecology. Karanth et al. (2003) provided a critical review about the census-based paradigm. They emphasize the importance of proposing alternative approaches concerning the problems of spatial sampling and observability. Bayliss (1987) related some of the variables that may affect the consistent detection of animals on a survey as observer experience, distance to animal, noise, animal behavior, weather and cover habitat.

Observability refers to the typical inability to detect and count all animals from the surveyed population, making observers to estimate the underlying detection probabilities from sample counts (Lancia et al. 1996, Thompson et al. 1998, Williams et al. 2002, Karanth et al. 2003). As Karanth et al. (2003) have pointed out, it has been generally assumed that the estimated proportion of the total area actually by survey (a) and the estimated proportion of counted animals in the surveyed area (p) is equal to 1, or that all animals are detected on the sampled area, assuming that a true animal census is carried out.

Capybaras (Hydrochoerus hydrochaeris), the largest grazing herbivore widely distributed in South America (Ojasti 1973, Azcárate 1980, Nowak & Paradiso 1991, Eisenberg & Redford 1999), have been considered as a potential plague in east central region of the State of São Paulo, Brazil, due to their higher population density in anthropogenic habitats (Verdade & Ferraz 2006) usually associated with crop damage (Ferraz et al. 2003), and the spread of Brazilian Spotted Fever (Labruna et al. 2004). Their high reproductive capacity, generalist food habits, and low habitat requirements are some aspects of the species biology that could contribute for their success in anthropogenic landscapes. In addition, the Brazilian hunting prohibition (Federal Law No. 5.197 from January 1967), the great food availability provided by the growth of cultivated lands, and the predators decline due to habitat loss (Costa et al. 2005) are the possible causes of the species population booming in these areas (Verdade & Ferraz 2006).

Capybaras are usually surveyed or monitored in many open flat habitats during their foraging activity by terrestrial (Ojasti 1973, Azcárate 1980, Cordero & Ojasti 1981, Macdonald 1981, Schaller & Crawshaw 1981, Herrera 1986, Jorgenson 1986, Alho et al. 1989, Herrera & Macdonald 1989, Verdade & Ferraz 2006) and aerial direct counts (Mourão et al. 1994; Mourão & Campos 1995, Mourão & Magnusson 1997). Correction procedures (as formerly suggested by Caughley 1977) are recommended in circumstances where the total population is unknown and its estimate is either needed or desired. However, it is likely that capybaras are less detectable in forest than in open areas. This problem could be more relevant in areas with high hunting pressure where capybaras forage at night time (Ojasti 1973, Verdade 1996).

This study aims to estimate capybara's detectability by an abundance index in forested habitats. This information might be useful for management purposes of the species for development correction factor in the counts of capybaras.


Material and Methods

Four study sites located in Campinas, east central region of the State of São Paulo, Brazil, were selected for this study (Figure 1). The study sites were carefully selected in order to provide accurate estimates of the true population size. This was only possible because all of them were fenced areas (i.e., no immigration) and the number of adults was previously known by the local offices. The observer only needed to check the births and deaths occurred in each survey. Habitats description is listed in Table 1.

Sampling surveys of capybara population was obtained by direct count (abundance index) and also by complete count (census) in 26 days on a weekly basis from March to September 2002. First, the observer just counted visible animals (direct count) while walking around the main water body in the study sites. Afterwards, he counted all animals he could find in the area (complete count). So the observer could have both estimates at the same time: the number of animals counted and the number of animals present in the area (true population size).

Detectability index (b), also called detection probability and observability, was estimated according to Lancia et al. (1996), Thompson et al. (1998) and Williams et al. (2002) as follows: E(Ci) = biNi, where expected value of count C in time i is equal to the detectability (b) in time i multiplied by the true population size N in time i. The probability of seeing or catching an animal (b) will generally be less than 1. As the detectability index is the chance of confirming the occurrence of an animal within some defined space and time (Thompson et al. 1998), this index could be used as observer error estimative in the estimates of animal population size.


Results and Discussion

The average detectability index of capybaras in forested habitats was 0.63 ± 0.32 (ranging from 0.31 to 0.95) for a single observer (Table 1). This means that the observer could be able to detect from 31% to 95% of the true capybara population in such conditions.

Thompson et al. (1998) pointed out that to be unbiased detectability index ought to assume that there is no influence of any factors such as weather, individual behavior, age and so on. As the observer and also the methodology of capybaras counting were the same during the whole study period we assume that the variability in the detectability index among habitats was due to the presence of more or less dense vegetation. The study sites Taquaral and Unicamp2 were habitats with no dense vegetation and large open areas where animals could be easily detected by the observer. Also, they were areas intensively used by people for leisure activities which means that animals could be more used to human presence. This could explain the higher detectability index by the observer in watching capybaras in these habitats.

Only 31.6% of the total population could be explained by the direct count considering all the study sites (p<0.001), described by the following equation: Total population survey = 6.992 + 0.7635(sampling population survey). In Taquaral, 35.3% of the capybara total population could be explained by the direct count (p < 0.001), described by the following equation: Total population survey= 7.150 + 0.6541(sampling population survey). The others sites had no significant relationship in the regression models between total and sampling population surveys. The weak relationship between total and sampling population surveys could be explained by the low variability of the total population numbers and the high variability of the abundance index in all study sites.

Despite the fact that our results are still preliminary they emphasize the importance and necessity of estimating observer error in animal counts for population size estimative, not only in forested habitats, but in open habitats too (Mourão et al. 1994). Verdade & Ferraz (2006) verified that distance between observers and capybaras can result in counting bias even in open habitats. When knowing observer error in counting animals it is possible to estimate population size in similar habitats. In addition, Bayliss (1987) recommend rigidly standardized survey procedures for minimizing sampling error.

In management programs, a harvest can be controlled either by placing a quota on offtake or by controlling harvesting effort (Caughley & Sinclair 1994). The control of harvest by quotas could not work properly with populations that fluctuate along time or those whose estimate is imprecise like capybaras in forested habitats. For those, controlling harvesting effort (as suggested by Caughley & Sinclair 1994) could be more effective.



The present study has been supported by Fundação de Estudos Agrários Luiz de Queiroz (FEALQ) and Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP).



ALHO, C.J.R., CAMPOS, Z.M. & GONÇALVES, H.C. 1989. Ecology, social behavior and management of the capybara in the Pantanal of Brazil. In Advances in Neotropical Mammalogy (K.H. Redford & J.F. Eisenberg, eds). Sandhill Crane, Gainesville, p.163-194.         [ Links ]

AZCÁRATE, T. de. 1980. Sociobiologia y manejo del capibara (Hydrochoerus hydrochaeris). Donana Acta Vert. 7(6):1-228.         [ Links ]

BAYLISS, P. 1987. Survey methods and monitoring within crocodile management programmes. In Wildlife Management: Crocodiles and Alligators (G.J.W. Webb, S.C. Manolis & P.J. Whitehead, eds). Surrey Beatty & Sons, Chippping Norton, NSW, Australia. p.157-175.         [ Links ]

CAUGHLEY, G. 1977. Analysis of vertebrate populations. John Wiley, London.         [ Links ]

CAUGHLEY, G. & SINCLAIR, A.R.E. 1994. Wildlife ecology and management. Blackwell Science Publications, Boston.         [ Links ]

CORDERO, R.G.A. & OJASTI, J. 1981. Comparison of capybara populations of open and forested habitats. J. Wild. Mngt. 45(1):267-271.         [ Links ]

COSTA, L.P., LEITE, Y.L.R., MENDES, S.L. & DITCHFIELD, A.D. 2005. Mammal conservation in Brazil. Conserv. Biol. 19(3):672-679.         [ Links ]

EISENBERG, J.F. & REDFORD, K.H. 1999. Mammals of the Neotropics: The Central Neotropics. v.2.University of Chicago, Chicago.         [ Links ]

FERRAZ, K.M.P.M.B., LECHEVALIER, M., COUTO, H.T.Z. & VERDADE, L.M. 2003. Damage caused by capybaras in a corn field, São Paulo, Brasil. Sc. Agric. 60(1):191-194.         [ Links ]

HERRERA, E.A. 1986. The behavioural ecology of capybara, Hydrochoerus hydrochaeris. Ph.D. dissertation, University of Oxford, Oxford.         [ Links ]

HERRERA, E.A. & MACDONALD, D.W. 1989. Resource utilization and territoriality group-living Capybaras (Hydrochoerus hydrochaeris). J. Anim. Ecol. 58:667-679.         [ Links ]

JORGENSON, J.P. 1986. Notes on the ecology and behavior of capybaras in northeastern Colômbia. Vida Silv. Neotr. 1(1):31-40.         [ Links ]

KARANTH, K.U., NICHOLS, J.D., SEIDENSTICKER, J., DINERSTEIN, E., SMITH, J.L.D., MCDOUGAL, C., JOHNSINGH, A.J.T., CHUNDAWAT, R.S. & THAPAR, V. 2003. Science deficiency in conservation practice: the monitoring of tiger populations in India. Anim. Cons. 6:141-146.         [ Links ]

LANCIA, R.A., NICHOLS, J.D. & POLLOCK, K.H. 1996. Estimating the numbers of animals in wildlife populations. In Research and management techniques for wildlife and habitats. (T.A. Bookhout, ed.). 5 ed. The Wildlife Society, Bethesda, p.215-253.         [ Links ]

MACDONALD, D.W. 1981. Dwindling resources and the social behaviour of capybara, (Hydrochoerus hydrochaeris) (Mammalia). J. Zool. Lond. 194:371-391.         [ Links ]

MOURÃO, G., BAYLISS, P., COUTINHO, M., ABERCROMBIE, C.L. & ARRUDA, A. 1994. Test of an aerial survey for caiman and other wildlife in the Pantanal, Brazil. Wild. Soc. Bull. 22:50-56.         [ Links ]

MOURÃO, G.M. & CAMPOS, Z. 1995. Survey of broad-snouted caiman (Caiman latirostris), marsh deer (Blastocerus dichotomus) and capybara (Hydrochaeris hydrochaeris) in the area to be inundated by Porto Primavera dam, Brazil. Biol. Conserv. 73:27-31.         [ Links ]

MOURÃO, G.M. & MAGNUSSON, W. 1997. Uso de levantamentos aéreos para o manejo de populações silvestres. In Manejo e Conservação de Vida Silvestre no Brasil (C. Valladares-Padua & R.E. Bodmer, eds). Sociedade Civil Mamirauá, p.23-33.         [ Links ]

NOWAK, R.M. & PARADISO, J.L. 1991. Walker's Mammals of the World. 5. ed. Johns Hopkins, London.         [ Links ]

OJASTI, J. 1973. Estudio Biológico del Chigüire o Capibara. Fondo Nacional de Investigaciones Agropecuarias, Caracas.         [ Links ]

SCHALLER, G.S. 1983. Mammals and their biomass on a Brazilian ranch. Arq. Zool. 31(1):1-36.         [ Links ]

VERDADE, L.M. 1996. The influence of hunting pressure on the social behavior of vertebrates. Brazilian J. Biol. 56(1):1-13.         [ Links ]

VERDADE, L.M. & FERRAZ, K.M.P.M.B. 2006. Capybaras on an anthropogenic habitat in Southeastern Brazil. Braz. J. Biol. 66(1b): 371 - 378.         [ Links ]

THOMPSON, W.L., WHITE, G.C. & GOWAN, C. 1998. Monitoring Vertebrate Populations. Academic Press, São Diego.         [ Links ]

WILLIAMS, B.K., NICHOLS, J.D. & CONROY, M.J. 2002. Analysis of Management o Animal Populations. Academic Press, San Diego.         [ Links ]



Date Received 08/08/2005
Revised 10/15/2005
Accepted 01/26/2006



ISSN 1676-0611

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License