Characteristics of the population of dogs and cats in Brazil

Article history The precise estimation of the domestic animal population is important for effective health planning. The objective was to verify the applicability in Brazil of the method proposed by the World Health Organization (WHO), to estimate the animal population of developing countries, comparing it with data published by the Brazilian Institute of Geography and Statistics (IBGE). From the questionnaire prepared by IBGE, specific questions were selected to be submitted for statistical evaluation. The results indicate a population larger than that estimated by the WHO, there is still a low vaccination coverage, and campaigns should be intensified in cats and in rural areas. Rural areas have larger numbers of cats and dogs per household, and the concentration of animals in the cities is higher in neighborhoods consisting of houses than apartments. The dog is the companion animal of choice of Brazilians. Received 24 September 2018 Accepted 12 January 2019


INTRODUCTION
The relationship of human beings with dogs and cats is very close and intense, and this relationship has an impact on the health of people and animals (VIEIRA et al., 2006). The inadequate breeding of animals and lack of knowledge about responsible ownership, associated with a low level of education, and lack of legislation, alters the population growth patterns of dogs and cats, which directly affects the well-being of all people concerned, and favors the transmission of diseases (LIMA; LUNA, 2012).
Although contact with animals carries risks, the frequency of most zoonotic diseases can be reduced or even eliminated, through population management practices and public policies aimed at veterinary care, restriction of movement, selective breeding, responsible legislation, and education of owners, making animal care a positive experience (CANATTO et al., 2012).
For planning these actions, it is of fundamental importance to know the size of the canine and feline population, and how it is geographically distributed (BELO;SILVA, 2015). In addition, population management programs should include, among other measures, a situation diagnosis, such as population estimates, population dynamics, and human attitudes and behaviors towards animals (GARCIA; CALDERÓN;FERREIRA, 2012). The analysis of these populations and their characteristics will be beneficial for the health and welfare of humans and non-humans in the Brazilian society (BIONDO et al., 2015).
According to Reichmann et al., (2000) several methods are proposed for estimating the canine and feline populations, such as animal census, sampling, questionnaires, and records from various sources. Given the impracticability of performing an animal census in several localities, the adoption of a methodology based on human population indicators has been shown to be more feasible. With these indicators, it is possible to evaluate programs of population control of dogs and cats, anti-rabies vaccination campaigns, and implementation of the necessary strategies in each region, according to their particularities (DIAS et al., 2004).
The WHO estimates that in emerging countries, the average proportion of dogs varies from 1:10 to 1:6, or about 10.0 to 16.7% of the human population (REICHMANN; NUNES, 1999). In Brazil, monitoring of the animal population is still based only on estimates of the human population, or on sample surveys.
Consequently, the objective of this study was to verify the applicability in Brazil of the method proposed by the World Health Organization (WHO), to estimate animal population of developing countries, comparing it with data published by the Brazilian Institute of Geography and Statistics (IBGE) in the National Health Survey (PNS).

MATERIAL AND METHODS
The PNS was produced by the IBGE nationwide in 2013, on the health situation and lifestyles of the Brazilian population, and for the first time in the history of Brazil the health survey questionnaire contained questions about companion animals. From this questionnaire, specific questions were selected to be submitted for statistical evaluation, to obtain the vaccine coverage, and the means of animal:person, animal:household, and dog:cat. The analysis of the microdata was conducted using the SAS 9.3 software, after access to the database obtained from the survey conducted by IBGE. According to the type of table, comparison between means, analysis of variance or independence tests were applied to verify the significance.
To determine the results of ratio cat and dog per person, the absolute number of animals per state and by type of census situation (urban or rural area) was found. The population quantification was obtained through a table produced in the IBGE Automatic Recovery System (SIDRA), based on the National Household Sample Survey (PNAD), and population projection for the year 2013, the year that PNS was conducted.
Regarding the vaccination coverage, from the absolute number of cats and dogs and the number of animals vaccinated against rabies in the last 12 months, the percentage of cats, dogs, and both vaccinated for each state was calculated, and by type of census situation. It is important to note that in the questionnaire, for the question about vaccination, the respondent was offered three types of responses: 1. Yes, all; 2. No, not all; 3. None of them. For the calculation of vaccination coverage, only the number of those who answered "Yes, all" was used, hence presenting a diagnosis that was as conservative as possible, with the minimum number of reliably vaccinated animals.
In relation to household information, the SAS shows the mean animal:household, weighted by the factor of sample expansion, the number of households with the presence and absence of animals, and the total number of households, according to the state, type of census situation, and household type (house or apartment). For the percentage that represents the "yes" within each unit, a rule of 3 was performed.

RESULTS
From the analysis performed, graphs and tables were elaborated, from which the following results can be extracted. Figure 1 shows the mean number of persons per cat, and per dog, considering each state, and Figure 2 shows the mean number of persons per cat and per dog in Brazil, by type of census situation. Note that the lower the mean, the greater the number of animals.   Considering cases where all animals in the household were reported to be vaccinated, divided by type of census situation, resulted in an observed total of 13,090,480 vaccinated cats, of which 9,654,187 were in urban areas, and 3,436,293 in rural areas, and 39,221,001 vaccinated dogs, 31,316,322 in urban areas, and 7,904,680 in rural areas. Figure 3 shows the vaccination coverage of cats and dogs in urban and rural areas.

DISCUSSION
From the microdata, there were estimated to be 22,070,633 cats and 52,198,324 dogs in Brazil, with a mean ratio of cat:person of 1:9.12, and dog:person of 1:3.85. The state of RS has the highest number of dogs per person (1:2.11) and the second largest number of cats per person (1:5.22), which justifies the actions that this state has been taking in the area of public policies for animals, such as the creation of a Special Secretariat for Animal Rights (SEDA) in the city of Porto Alegre in 2011, with the aim of reducing animal mistreatment, controlling the population growth of dogs and cats, significantly reducing the abandoned animal population, and promote responsible animal ownership (PORTO ALEGRE, 2011).
PR comes in second place in the mean ratio of dog:person (1:2.48), a high index that possibly contributed to the Municipality of Curitiba creating the Defense and Animal Protection Network, a program that involves several public agencies, private initiatives, and the third sector in the search for better living conditions for the city's domestic animals (SILVA, 2009). SP is the federal unit with the largest animal population, with 3,866,973 cats and 10,558,406 dogs. This substantial quantity is accompanied by cutting-edge actions, and SP currently has the most advanced animal protection legislation in the country, with a wellestablished birth control policy in several municipalities. In June 2018 a Policy and a State System for the Defense of Domestic Animals weres created, with the objective of promoting joint action between the state and municipalities for the protection of domestic animals (SÃO PAULO, 2018). The municipality of São Paulo also has two public veterinary hospitals, fully financed with public resources (SÃO PAULO, 2009), a pioneer initiative that other states are also trying to implement.
Regarding the type of census situation, although in absolute numbers the population of cats and dogs is higher in urban areas (14,629,322 and 38,561,339) than in rural areas (7,441,311 and 13,636,986), it is observed that proportionally the concentration of cats and dogs per person is higher in rural areas (1:4.12 and 1:2.25) than in urban areas (1:11.67 and 1:4.43). This may be because public policies for animals, information on responsible ownership and castration campaigns are scarcer in rural areas.
The WHO estimates that, in emerging countries, the mean proportion of dogs varies from 1:10 to 1:6 of the human population (REICHMANN; NUNES, 1999). When comparing the population estimate of dogs using the WHO methodology with the results found in this research, it is observed that in total, and in all the federal states, except for AL and RN, the number of dogs was higher than that indicated by WHO, demonstrating that Number of Dogs per Cats the population of dogs is markedly higher than expected, a fact relevant to health planning.
Other studies with other methods and in different localities had also identified these underestimated values, as described in the municipality of Taboão da Serra -SP, with a person:cat ratio of 30.57, and a person:dog ratio of 5.14 (DIAS et al., 2004); in the interior of São Paulo, with a ratio 16.4 for cats and 4.0 for dogs (ALVES et al., 2005); in the municipality of São Paulo, with 29.49 for cats and 7.28 for dogs (MAGNABOSCO, 2006) and 19.33 for cats and 4.34 for dogs (CANATTO et al., 2012); in Curitiba -PR, with 86.38 for cats and 13.05 for dogs (SERAFINI et al., 2008); in Cacoal -RO, with 11.2 for cats and 4.5 for dogs (BELO; SILVA, 2015); and in São José dos Pinhais -PR, with 15.32 for cats and 2.47 for dogs (CATAPAN et al. , 2015).
An underestimation of the animal population may jeopardize vaccination programs for rabies and control of other zoonoses, since if the canine estimate is based solely on WHO criteria, the number of animals needed to be vaccinated may be underestimated and thus vaccination coverage lower than desired (DIAS et al., 2004), leading to a misunderstanding of the campaign results. More accurate population estimates, through a continuous census of domestic animals, would allow the animal population to be safely known and its changes tracked.
The results indicate a rate of 75.13% of dog vaccination, which is a satisfactory coverage according to WHO recommendations. The WHO has established the goal of at least 75% of the estimated canine population (REICHMANN; PINTO; NUNES, 1999). However, furthering this analysis by state, it is found that RO, AC, AM, RR, PA, AP, MA, PI, PE, BA, PR, SC, and RS did not reach the minimum required index. Note that this result includes practically the entire North Region, excluding TO, and the entire South Region, which is very worrying, since the south is also among the highest rates of dogs per household in the country. The DF was the only state with vaccine coverage rates above 75% for cats (75.28%) and for dogs (89.00%).
Regarding the type of census situation, it is observed that the vaccination coverage for rabies is higher for cats (65.99%) and dogs (81.21%) in urban areas, showing that the vaccination campaigns have not reached rural areas (cats 46.17% and dogs 57.96%) with the same efficiency, leaving these regions more vulnerable. An aggravating factor for the results found, is that in rural areas, dogs may be easier targets of rabies infection, transmitted by other domestic dogs, chiropterans, or other wild animals (REICHMANN; PINTO; NUNES, 1999), so these regions need intensified actions. Among the several causes of low vaccination coverage, the logistic difficulty of vaccine distribution is a fact that may explain the low results in the North Region and in rural areas. The DF has a small territory, crossed by several routes, which certainly contributed to the good indexes.
A more detailed analysis by species showed that although the overall canine index reached the target (75.13%), the overall feline coverage was only 59.31%, and in all states the feline coverage was also smaller than that of dogs. It can be inferred that owners find it more difficult to take their cats to the vaccination posts than their dogs. In addition, there is a culture that cats need less veterinary care, or their needs are less well known, and many live with less restriction of movement, making it difficult to locate and capture them (GARCIA, 2009).
Laboratory results regarding the diagnosis of rabies in the state of SP have prompted an increase in the percentage of positive cases in felines (ALVES et al., 2005), probably because rabies control measures are predominantly directed at dogs, rather than at cats and wild animals (REICHMANN; NUNES, 1999). A more careful evaluation of the feline population is suggested, as well as its epidemiological importance in the transmission of rabies, and the vaccination of cats is recommended and necessary.
In the survey carried out by IBGE, there were estimated to be 65,195,293 households in Brazil, with 17.65% having a cat and 44.26% a dog. Specifically, in SP state, the results were 13.44% and 43.37%, respectively. Comparatively, other authors found similar results, such as Alves et al. (2005) in the interior of São Paulo, who evaluated 20,958 households and found that 12.6% had cats and 52.6% had dogs, similar to the results described by Canatto et al. (2012).
Of the households interviewed in the municipality of Pinhais (PR) by Martins and coworkers (2013) On the other hand, the high number of animals found in a single household is noteworthy. Within the states, the variation is from 11 to 51 cats and from 7 to 80 dogs, with 51 cats in one house in a rural area in RS, and 80 dogs in one house in an urban area in the Federal District (DF). In apartments a maximum of 10 cats and 21 dogs were found. The promotion of well-being in conditions of high agglomeration of animals is not feasible, and the health of the residents can also be compromised (TEIXEIRA; SILVA; SOARES, 2016). It is important that public managers be aware of and have mapped cases of animal hoarding in their region, so that these residents and animals can be adequately assisted (FILHO et al., 2013).
In the same way that the number of animals per person was higher in rural areas, the result is repeated for the mean number of animals per household, with 0.26 cats and 0.69 dogs per household in urban areas and 0. Concerning the presence of animals in apartments, only 5.75% had a cat and 18.26% a dog, but in houses, 19.28% had a cat and 47.88% a dog. The differences in the number of dogs and cats in neighborhoods with a predominance of apartments or houses, emphasizes the importance of taking into account this information for population control and vaccination programs (SERAFINI et al., 2008).
The expansion and enrichment of cities, and demographic changes are expected to rapidly increase the number of felines in many countries. In the United States, France, and Germany, the population of cats is already larger than that of dogs. This is because dogs are better adapted to conditions which are becoming scarcer, with spacious homes, large families, and time available for the care that the species demands. On the other hand, cats do their own cleaning, use sandboxes, and adapt well to apartments, as they take advantage of vertical space. This indicator of progress based on the domestic feline population is called the Big Cat, alluding to the Big Mac index, which measures the valuation of each country's currencies (TEIXEIRA, 2013). But, Brazil presented a proportion of dogs to cats of 2.36 and in all states the preference was still for dogs. The DF ranked first in the mean ratio of dog:cat (4.15), followed by PR (3.54), and MG (3.53). On the other hand, CE state showed the lowest preference for dogs, obtaining a result of 1.13, followed by PI (1.31), and MA (1.33).
Concerning the type of census situation, urban areas had a higher number of dogs (2.63) than the number of cats, compared to rural areas (1.83), which can be explained by the fact that the dog is still the companion animal of choice of large urban centers. The preference for dogs was higher in apartments (2.80) than in houses (2.35), similar to that described by Biondo; Martins; Ferreira (2014) (Houses 6.82:1 and apartments 6.61:1), emphasizing that the household verticalization of urban areas in Brazil, despite the common belief and pattern of change in other countries, did not affect owners' preference for dogs compared to cats.
It is important to emphasize that the PNS was not designed to achieve the objectives of this research, and may generate some limitations, however this does not disqualify the results obtained.

CONCLUSION
The methodology proposed by the WHO to estimate the animal population of developing countries presents smaller population numbers in comparison to the IBGE data, since it was found that the ratio between number of animals and people cannot be applied equally to all states, nor does it equate dogs and cats in this proportion. The national rabies vaccination coverage of dogs does not meet the minimum recommended by WHO, in some states and in rural areas.
Rural areas have larger numbers of cats and dogs per person and per household, just as the concentration of animals in the cities is higher in neighborhoods of houses, than in neighborhoods where apartments predominate. Even with one of the largest populations of domestic animals in the world, few states averaged more than one dog per household, and in no case was there an average of more than one cat per household, thus making it possible for households to absorb animals, a situation that favors the adoption campaigns of animals. Finally, the dog is the companion animal of choice of Brazilians.