Hunting of Mammal Species in Protected Areas of the Southern Bahian Atlantic Forest, Brazil

Hunting of mammal species in protected areas of the southern Bahian Atlantic Forest, Brazil

L UCIANA C. CASTILHO,KRISTEL M. DE V LEESCHOUWER,E.J.MILNER-GULLAND and A LEXANDRE S CHIAVETTI

Abstract To investigate the practice of hunting by local peo- social and cultural traditions (Bennett & Robinson, ; ple in the southern Bahia region of Brazil and provide infor- Secretariat of the Convention on Biological Diversity, mation to support the implementation of the National Action ). However, wildlife harvest has reached unsustainable le- Plan for Conservation of the Central Atlantic Forest vels in many places because of the increasing human popu- Mammals, we conducted  interviews with residents of lation and demand for wild meat (Peres, ; Bennett et al., three protected areas and a buffer zone. Thirty-seven percent ), improved hunting technologies and increased access of respondents stated that they had captured an animal op- to forests (Bennett et al., ;Fa&Brown,). Besides portunistically, % hunted actively and %didnothunt. causing species declines or extinctions (Cullen et al., ; The major motivation for hunting was consumption but peo- Corlett, ; Peres & Palacios, ), overhunting can affect ple also hunted for medicinal purposes, recreation and retali- the ecological functionality of ecosystems (Cullen et al., ; ation. The most hunted and consumed species were the paca Wright et al., ; Peres et al., ), as well as local commu- Cuniculus paca, the nine-banded armadillo Dasypus novem- nities that depend on the consumption of wild meat for sub- cinctus and the collared peccary Pecari tajacu; threatened spe- sistence (Milner-Gulland et al., ). The importance of cies were rarely hunted. Opinions varied on whether wildlife hunting in household economies varies depending on socio- was declining or increasing; declines were generally attributed economic factors such as level of education (Nielsen & to hunting. Our findings suggest there is illegal hunting for Meilby, ), wealth, household size (Foerster et al., ), consumption in and around protected areas of the region. livestock ownership (Loibooki et al., )andage(Melo Management efforts should prioritize fairness in the expropri- et al., ). Unsustainable hunting is a major concern ation process for people who must be relocated, and adopt an both for wildlife conservation and with respect to the well- approach to wildlife management that involves residents being of the people that subsist on wild meat (Redford, living around the protected areas, and considers their needs. ; Nasi et al., ). In Brazil, hunting is illegal (Federal Law /—Law of Keywords Brazil, hunting, protected area, randomized re- Environmental Crimes and Decree /), and penal- sponse technique, semi-structured interviews, threatened ties are more severe for hunting inside protected areas and mammals, wild meat, wildlife conservation for the hunting of threatened species (Art.  Federal Law Supplementary material for this article is available at /). Nonetheless, hunting is widespread (Carvalho & https://doi.org/./S Morato, ; El Bizri et al., ) and has been cited as a major threat to wildlife in the protected areas of the Atlantic Forest (Chiarello, ; Canale et al., ;  Introduction Schiavetti et al., ), the Amazon (Carvalho & Pezzuti, ) and other regions in the north-east of the country unting in tropical forests provides nutrition and in- (Fernandes-Ferreira et al., , ). Subsistence is the Hcome for local communities and is deeply rooted in main reason for hunting, particularly in indigenous and traditional communities (Peres & Nascimento, ; Hanazaki et al., ; Barbosa et al., ; Minzenberg & LUCIANA C. CASTILHO* (Corresponding author) Graduate Program in Ecology Wallace, ). However, recreational activity and trade and Biodiversity Conservation, State University of Santa Cruz, Rodovia Ilhéus-Itabuna, km16 Salobrinho, CEP 45.662-090, Ilhéus, Bahia, Brazil are also motivations for hunting (Chiarello, ; Alves E-mail [email protected] et al., ; Fernandes-Ferreira et al., ; El Bizri et al., KRISTEL M. DE VLEESCHOUWER Centre for Research and Conservation, Royal ). Threatened mammals have been captured for the Zoological Society of Antwerp, Antwerp, Belgium pet trade (Fernandes-Ferreira et al., ; Nascimento E. J. MILNER-GULLAND Department of Life Sciences, Imperial College London, et al., ) or killed for consumption (Castilho et al., ; and Department of Zoology, University of Oxford, Oxford, UK Melo et al., ; Morcatty & Valsecchi, ), traditional ALEXANDRE SCHIAVETTI Department of Agricultural and Environmental Sciences,   State University of Santa Cruz, Ilhéus, Bahia, Brazil medicine (Alves, ; Ribeiro et al., ), or in retaliation for wildlife-related damage (Carvalho & Morato, ). *Also at: Centre for Research and Conservation, Royal Zoological Society of Antwerp, Antwerp, Belgium In an attempt to mitigate threats to threatened species in Received  February . Revision requested  April . Brazil, National Action Plans have been developed to iden- Accepted  July . First published online  November . tify and prioritize conservation actions (ICMBio, ). The

Oryx, 2019, 53(4), 687–697 © 2017 Fauna & Flora International doi:10.1017/S0030605317001247 Downloaded from https://www.cambridge.org/core. IP address: 170.106.202.226, on 30 Sep 2021 at 17:31:56, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0030605317001247 688 L. C. Castilho et al.

Atlantic Forest receives special attention because it is con- southern Bahia. There are c.  private properties in the sidered to be a hotspot for biodiversity (Myers et al., four areas. ). Major threats to species of the Atlantic Forest are Una Biological Reserve (IUCN category Ia) comprises the increasing human population and associated activities, one of the largest remnants of Atlantic Forest in southern such as deforestation, hunting and illegal wildlife trade Bahia (Schroth et al., ). It was established in (ICMBio, ). In  a National Action Plan was created December  and expanded from , to , ha in to protect threatened mammals of the Central Atlantic December . At the same time, in December , the Forest, with the aim of increasing the population viability Reserve’s original buffer zone was designated Una Wildlife of its target species and improving the quality of their habi- Refuge (, ha, IUCN category III). Serra das Lontras tats (ICMBio, ). For the southern Bahia region one of National Park (, ha, IUCN category II) was established the main goals of this Action Plan is to decrease hunting in June  and comprises one of the last remnants of mon- pressure on target species within key conservation areas tane Atlantic Forest in north-eastern Brazil (Schroth et al., (ICMBio, ). ). The Park has a buffer zone of , ha. The law of Our research aims to improve understanding of the prac- the National System of Conservation Units (Art.  and  tice of hunting by local people in the southern Bahia region. Federal Law ./) states that all privately owned Recording information on sensitive and illegal behaviour areas inside Una Biological Reserve and Serra das Lontras such as hunting inside and around protected areas is challen- National Park should be expropriated, whereas private ging, especially when residents are aware of the illegality of land ownership is allowed within Una Wildlife Refuge, as the activity. Novel approaches have emerged to obtain long as land use is compatible with the conservation more accurate information about such sensitive behaviours goals of the Reserve. If this condition is not met, this land (Nuno & St John, ). We tested a relatively recently must also be expropriated (MMA, ). According to developed method of indirect questioning for eliciting sen- the Chico Mendes Institute for Biodiversity Conservation sitive information (the randomized response technique), as (ICMBio), which manages all federal protected areas, expro- a means of obtaining an estimate of the prevalence of illegal priation of private properties within the boundaries of Serra hunting of species of concern (Solomon et al., ; St. John das Lontras National Park has not yet been initiated. In the et al., ; Razafimanahaka et al., ). This was the longer established Una Biological Reserve, % of properties first use of this technique in Brazil in the context of wildlife have been expropriated, but no expropriations have conservation, and we explore its suitability for use in this occurred in the Wildlife Refuge. context. A number of studies have explored the prevalence of, and motivations for, hunting by rural people; for example, Nuno Methods et al. () used an indirect questioning method to explore bushmeat hunting in the Serengeti, Tanzania, as did Target species Harrison et al. () in the Bwindi area in Uganda. We focused on mammal species that occur in southern Adding to this literature, our study was the first integrated Bahia, especially the threatened species listed in the assessment of the factors affecting hunting of threatened National Action Plan for the Conservation of Central mammals in the Bahia region. We explored the relationship Atlantic Forest Mammals as well as mammals that are com- between socio-economic factors and hunting prevalence to monly hunted by local communities in the Atlantic Forest understand whether particular sectors of society are more or region (Table ). less dependent on hunting. This type of information not only supports the implementation of Brazil’s National  Action Plan (ICMBio, ) but also adds new insights to Data collection improve general understanding of the factors affecting hunting behaviour among people living in and around Data were collected during October –May .We protected areas. conducted  interviews within the three protected areas and the buffer zone, covering % of the properties located within the four areas. These comprised direct semi- Study area structured interviews with rural residents in all four areas, and interviews using the randomized response technique The research was conducted in three protected areas (Una with a different sample of respondents in Una Wildlife Biological Reserve, Una Wildlife Refuge and Serra das Refuge and the buffer zone. When possible, interviews Lontras National Park) and a buffer zone (Fig. ). The pro- were performed in the presence of a local field assistant tected areas cover a total area of , ha and are located who introduced us and helped us to connect with the re- mostly in the municipalities of Una and Arataca, in spondents. We assured respondents of the confidentiality

FIG. 1 The study area comprising three protected areas and a buffer zone in southern Bahia, Brazil.

of their responses, and all interviews were conducted with residence, number of people living at home, receipt of govern- their consent (State University of Santa Cruz—Ethics ment assistance (Bolsa Família, a programme of supplemen- Committee CAAE: ...). Authorization tary income offered to low-income families) and presence of to perform research inside protected areas was provided livestock (chickens, pigs, sheep, ducks and geese). Previous re- by the Chico Mendes Institute for Biodiversity search findings on the influence of socio-economic factors on Conservation (SISBIO-ICMBio: -). hunting activities are described in Supplementary Table S. To select a random sample of interviewees, we obtained data from the Chico Mendes Institute containing georefer- Part II: Respondents’ perceptions of wildlife and hunting ac- enced locations of properties inside the protected areas and tivities in the villages. We used photographs of the target in part of the buffer zone. For both interview methods we species to verify recognition and perception of occurrence used a random number generator to choose properties to sam- by residents. ple from this list. However, in several cases selected properties were unoccupied or we were unable to locate the resident, and Part III: Information about hunting (whether the respond- therefore we also conducted opportunistic sampling, including ent hunted inside protected areas, their motivations, and additional properties to increase the sample size. On arrival at which species were hunted). If the respondent was a a property we interviewed the person responsible for it or any- woman we asked her to answer this part of the interview one who was available, including both landowners and work- on behalf of her partner or the household head, as appropri- ers. Only one person per property was interviewed. ate. The responses suggested that all members of a household had a good awareness of the hunting behaviour of the other household members. As we did not probe for a deep understanding of motivations and techniques, but fo- Direct interviews cused on factual information, we felt that the responses from We approached  respondents, of whom  agreed to be each type of respondent were representative of the house- interviewed. Of these,  were male and  were female. hold as a whole. For respondents who were answering on This high response rate was probably because our initial ap- behalf of their partners, their age, level of formal education proach focused on non-sensitive subjects. However, we and duration of residence were usually correlated with their sensed that most people were not comfortable talking partners’ characteristics, and the rest of the socio-economic about hunting, as they knew that this activity was illegal. variables were at the household level. We considered two In  interviews, respondents denied association with hunt- forms of hunting: opportunistic capture (an individual cap- ing but provided information that suggested an association. tures and kills a wild animal if they encounter it near their Despite this we assumed they were non-hunters. house or during work) and active hunting (pre-meditated The interview was organized into three parts hunts using traps, and active searching using firearms and (Supplementary Table S): dogs). Consequently, to verify the prevalence of hunting we asked two questions: () Do you hunt (actively) within Part I: Socio-economic information on respondents, including the property or nearby? () Do you capture a wild animal their gender, age, level of formal education, duration of if you find it opportunistically? We also showed

TABLE 1 Focal species in our study of hunting of mammals in pro- responses (Gavin et al., ). Studies have shown that esti- tected areas in southern Bahia, Brazil (Fig. ), with their conserva- mates of illegal behaviour based on this technique are higher tion status on the IUCN Red List, and whether or not they were compared to those based on direct questioning, and there- included in interviews using the randomized response technique. fore it is a useful technique for assessing sensitive behaviour Randomized such as illegal resource use (Solomon et al., ; St. John IUCN response et al., ; Razafimanahaka et al., ). Species status1 technique We used the randomized response technique to investi- Preferred species gate consumption of mammals, focusing on four species Paca Cuniculus paca2,3 LC Yes known to be preferred, and five threatened species, all of Nine-banded armadillo Dasypus LC Yes  2,3 which are illegal to hunt anywhere in Brazil (Table ). We novemcinctus also included chicken as a control. The approach was adapted Naked-tailed armadillo Cabassous LC No 2 from Solomon et al. (). Based on a pilot study, we used unicinctus ‘ ’ Collared peccary Pecari tajacu3,4 LC Yes an approach known as the two unrelated questions tech- Opossum Didelphis aurita2,5 LC No nique; one question is related to the sensitive behaviour Deer Mazama sp.3,4 LC Yes and the other question is related to a non-sensitive behav- National Action Plan6 species iour. Firstly, we asked the respondent to identify which of (ICMBio, 2010) nine wildlife species they recognized from photographs. Northern brown howler monkey CR No Then we separated out the recognized species, to be used Alouatta guariba guariba in the interview. For each species we showed two cards to Yellow-breasted capuchin Sapajus CR Yes the respondent, one with a picture of the animal and one xanthosternos with a picture of a coin, explaining the questions that each Unicolored tree-rat Phyllomys CR No unicolor card represented, and then placed them inside identical en- Golden-headed lion tamarin EN Yes velopes. Secondly, we gave a coin to the respondent to flip Leontopithecus chrysomelas before selecting an envelope. The respondent selected an en- Southern Bahian masked titi VU Yes velope and answered only yes or no to the question inside, Callicebus melanochir without telling us which card they were looking at. If they se- Maned sloth Bradypus torquatus VU Yes lected the photograph of an animal, they would answer yes if Thin-spined porcupine Chaetomys VU Yes they had eaten that animal in the previous  months, or no subspinosus Painted tree-rat Callistomys pictus VU No otherwise. If they selected the coin card, they would answer yes if they had got heads when they flipped the coin, and no  LC, Least Concern; VU, Vulnerable; EN, Endangered; CR, Critically for tails. The interviewer did not know which card the re- Endangered  spondent had chosen, ensuring anonymity. In addition, we Pereira & Schiavetti ()  Flesher & Laufer () obtained socio-economic information from respondents.  Cullen et al. () Of the  people approached for the randomized re-  Hanazaki et al. ()  sponse technique interviews (a non-overlapping sample National Action Plan for the Conservation of Central Atlantic Forest with the direct interview survey),  (%) did not wish to Mammals participate. Two of them refused because they did not understand the method and the others refused to participate photographs of threatened species to verify if respondents before the method was explained, suggesting the sensitivity had consumed these species. We had a lower number of re- of the subject was the main reason for refusal. For people sponses to these questions because this was the last part of who found it difficult to understand the method, we ex- the interview and we did not have enough time to ask these plained it carefully until they felt satisfied to participate. questions to all respondents:  respondents answered the Overall, we received  responses to the randomized re- questions related to the golden-headed lion tamarin sponse technique interviews. Leontopithecus chrysomelas and the southern Bahian masked titi Callicebus melanochir and  respondents answered the question related to the maned sloth Bradypus torquatus. Data analysis We used a generalized linear model (logistic analysis of covari- Randomized response technique ance) to verify whether hunting in a location (binary response) was affected by respondents’ demographic characteristics The randomized response technique was developed by (explanatory variables; Supplementary Table S). We used Warner () to eliminate evasive answers and response Pearson correlation coefficients to explore associations bias in sensitive questions, guaranteeing confidentiality of among the explanatory variables. As there was no strong

correlation between variables they were all included in the TABLE 2 Numbers of interview respondents who reported wild spe- model. We used ANOVA for model simplification to cies to be increasing (n = ) or decreasing (n = ) in number, or  find the minimal adequate model that also had the lowest among the most hunted species (n = ), in three protected areas   Akaike information criterion value. We used Pearson’s χ and a buffer zone in southern Bahia, Brazil (Fig. ). Threatened species are in bold. (contingency tables) tests to explore associations between respondents’ perceptions and locations, and simple logistic re- No. of reports gression to explore the relationship between age and Species Species Most respondents’ perceptions of changes in wildlife abundance. Species increasing decreasing hunted To estimate the proportion of people consuming the spe- Armadillo (Dasypus 23 67 cies investigated using the randomized response technique, novemcinctus, Cabassous we used the following equation (Fox & Tracy, ): unicinctus) Collared peccary 20 7 35 l −(1 − P)p Golden-headed lion X = , 11 P tamarin and to calculate the estimated variance for the population Marmoset Callithrix 7 surveyed we used: kuhlii Paca 6 6 99 l(1 − l) Capybara Hydrochoerus 64 3 Var(X)= nP2 hydrochaeris where X = estimated proportion of respondents consuming Crab-eating fox 4 λ ‘ ’ Cerdocyon thous the species, = recorded proportion of yes responses, Coati Nasua nasua 34 π   = known proportion of non-sensitive behaviour (= . , Southern Bahian masked 2 i.e. ﬂipping heads in the coin toss), P = probability of select- titi ing the sensitive question (= .) and n= number of respon- Opossum 34 3 dents. We performed regression modelling using a Deer Mazama sp. 1 18 5 customized link function (Keane, ) to explore associa- Agouti Dasyprocta 10 1 tions between the estimates of consumption obtained leporina Maned sloth 10 from the randomized response technique and socio- Collared anteater 17 3 economic variables. Tamandua tetradactyla We conducted fewer interviews in Una Biological Puma Puma concolor 13 Reserve and Serra das Lontras National Park because of dif- Kinkajou Potos flavus 12 ferences in numbers of residents and challenges in accessing Yellow-breasted 2 some properties. We therefore grouped the Reserve with capuchin Una Wildlife Refuge and the Park with the buffer zone, based on the proximity of the areas and the history of management actions for the protected areas. Socio-economic characteristics of respondents are described in Supplementary Fifty-four percent of residents said that overall wildlife abun-  Table S. dance was increasing compared to years previously, whereas % said that it was decreasing. Perceptions of change in abundance were unrelated to respondents’ age or location. Results Several species were reported by respondents to have disappeared or decreased in abundance, and most respondents  Perceptions of wildlife abundance and links to hunting ( %) cited overhunting as the main reason. The reasons given for increased abundance were rapid reproduction or a Among the targeted species (Table ), armadillos (Dasypus decrease in hunting of the species (e.g. the golden-headed lion novemcinctus and Cabassous unicinctus), collared peccaries tamarin). Respondents reported that pacas and armadillos Pecari tajacu and pacas Cuniculus paca were the most recog- were the most hunted species in the region (Table ). nized, and the ones respondents were most likely to say were Species believed by respondents to have disappeared more present in the study area. The northern brown howler mon- than  years previously were the lowland tapir Tapirus ter- key Alouatta guariba guariba and the painted tree-rat restris, the northern brown howler monkey, the white-lipped Callistomys pictus were the least frequently recognized spe- peccary Tayassu pecari, the northern muriqui Brachyteles hy- cies, suggesting their absence or lower occurrence in the poxanthus and the giant armadillo Priodontes maximus. area (Supplementary Table S). It was not possible to obtain Fifty-six percent of respondents perceived a reduction in information on the unicolored tree-rat Phyllomys unicolor be- hunting activities in their villages and surrounding areas in cause residents were unable to identify the species reliably. recent years and % said that hunting had not changed in

TABLE 3 Numbers of interview respondents in three protected areas and a buffer zone in southern Bahia, Brazil (Fig. ), who stated that they hunted in protected areas of the southern Bahian Atlantic Forest, the numbers of these who hunted opportunistically and actively, and the numbers who stated they did not hunt.

No. of respondents Una Wildlife Refuge Una Biological Serra das Lontras Buffer zone Total (n = 74) Reserve (n = 20) National Park (n = 11) (n = 64) (n = 169) Yes 31 12 9 37 89 Opportunistic 25 8 7 23 63 Active 6 4 2 14 26 No 43 8 2 27 80

prevalence. More respondents in Una Biological Reserve– they hunted, in response to a direct question, included Una Wildlife Refuge perceived a decrease in hunting com- only level of formal education (Table ; for details of models pared to respondents in Serra das Lontras National Park see Supplementary Table S); respondents educated to pri-  and the buffer zone (χ = .,df=,P=.). mary level hunted more than those with either no education or with other levels of education.

Prevalence of hunting and motivations

Thirty-seven percent of respondents said that they hunted Wild meat consumption opportunistically, % said they hunted actively and % Using the randomized response technique, the estimate of said they did not hunt. The highest level of active hunting whether respondents had consumed chicken in the previous was in the buffer zone (% of respondents said they were  months was not significantly different from %. active hunters; Table ). Active hunters used hunting dogs Estimates of consumption prevalence were non-zero for ar- (%), traps (%) and hunting platforms (%). The pri- madillos ( ± SE %), pacas ( ± SE %) and peccaries mary stated motivation for hunting was consumption ( ± SE %). For the other species, including all the threa- (% of responses). However, wild meat was an occasional tened species, the confidence intervals for consumption complement to diets rather than being consumed regularly. overlapped zero (Fig. ). Wide confidence intervals are a In general discussion many respondents stated that wild common issue with the randomized response technique, meat was essential for subsistence in the past, but not any- given the probabilistic nature of the method, requiring a more. Although most hunters said they killed because they large sample size. However, the confidence intervals give enjoyed a particular animal’s meat, some respondents said an upper bound on the likely prevalence of consumption they hunted out of necessity, because of a lack of animal pro- of these less-consumed species. tein. Most opportunistic hunters stated that they did not With respect to socio-economic factors affecting the es- consider themselves to be hunters. Besides hunting for con- timated proportion of consumption based on the rando- sumption, % of respondents who had hunted had used mized response technique, the data were sufficient for wild animals for medicinal purposes, % had hunted in re- generalized linear models for only three species (paca, nine- taliation for damage to crops or livestock and % said that banded armadillo and collared peccary), and only one had a hunting was a recreational activity (% among active hun- significant result: the estimated consumption of collared ters). According to respondents the species used most in peccary tended to be higher for older people (Table ; for de- popular medicine are pacas, porcupines (especially the tails of candidate models see Supplementary Table S). Bahia hairy dwarf porcupine Coendou insidiosus), the In the direct interviews, respondents cited several species black-and-white tegu Tupinambis merianae and armadillos. that were preferred for consumption (Table ). Few people We were unable to identify anyone involved in hunting for said that they consumed any of the threatened species; for commercial purposes. However, some respondents provided example, % of the  respondents who answered the ques- information about prices of wild meat, suggesting the exist- tion about whether they ate monkeys said they did not, and ence of an illegal local trade (Supplementary Table S). %of people said they did not eat sloths. The reasons threatened species were not eaten included taboos or Associations between hunting and socio-economic cultural norms, or because they were non-preferred species. factors Some respondents did not consume sloths and monkeys because they look like people. Others said they did not Of the initial set of seven explanatory variables, the minimal consume sloths because ‘they are like women’, meaning adequate model explaining whether someone stated that they have the same menstrual cycle as women. People also

TABLE 4 Minimal adequate model (logistic analysis of covariance) affected our ability to estimate consumption of the less showing effects of level of education on whether a respondent sta- widely consumed species. Other indirect questioning tech- ted they hunted (either actively or opportunistically). The baseline niques, such as the unmatched count technique (Nuno et al., level of education is high-school/college. ), which are easier for respondents to understand, Estimate SE z P(. |z|) would also be worth piloting in this landscape. (Intercept) −0.6931 0.4629 −1.497 0.1343 Education (none) 0.5390 0.5255 1.026 0.3051 Hunting in protected areas Education (primary) 1.3041 0.5436 2.399 0.0164* Education (middle-school) 0.5978 0.6365 0.939 0.3476 Our results suggest that mammals are hunted in and around *P , . protected areas in the southern Bahia region. Most people who hunted did so occasionally, and predominately to complement their diet, although also for medicinal reasons, recre- said they did not consume porcupines (including the thin- ation and retaliation. Consumption of wild meat is common spined porcupine Chaetomys subspinosus) because the meat in Brazil, both for subsistence (Peres, ; Hanazaki et al., is considered to be remosa, which means that it is not good ; Barbosa et al., ) and because rural residents appre- for consumption by sick people or pregnant women. ciate its flavour (Fernandes-Ferreira et al., ). We also ob- Consumption of these less desirable species was often said tained some indirect evidence of commercial hunting for the to be related to opportunistic situations, because of a need illegal wild meat trade. Our results suggest that the majority for protein and their ease of capture when encountered of hunting occurring in the area is opportunistic. People who opportunistically. capture wildlife opportunistically probably do not consider themselves to be hunters because they do not perceive hunt- Discussion ing for personal consumption as a profession but as a part of everyday life. Controlling this type of hunting may be a great- Although legal subsistence hunting by indigenous and trad- er challenge than controlling targeted hunting, as it offers an itional communities has been widely studied in Brazil unexpected opportunity for people to supplement their diet (Hanazaki et al., ; Pereira & Schiavetti, ; Barbosa (Gardner & Davies, ) and because it is also related to peo- et al., ), especially in the Amazon region (Peres, ; ples’ feelings and attitudes about their relationship with wild- Peres & Nascimento, ; Minzenberg & Wallace, ), life. Our finding that hunting occurs in protected areas illegal hunting in strictly protected areas has been poorly in- within the study area was supported by researchers from vestigated. People are aware that hunting is illegal, which other projects meeting hunters and finding traps and hunting makes research on the subject difficult (Alves & Souto, platforms in Una Biological Reserve. We also found evidence ). We found that using a variety of techniques was a of hunting (traps and dogs) during our fieldwork in all areas, valuable strategy for collecting information about illegal but more frequently in the buffer zone. hunting and to triangulate our results. Our respondents Respondents were split on whether hunting had in- were sometimes uncomfortable with the subject matter, creased or decreased and whether species abundance had but nonetheless through direct interviews we were able to changed. The perception of a decrease in hunting was higher obtain detailed information about hunting, and explore re- in Una Biological Reserve–Una Wildlife Refuge. The ab- spondents’ perceptions and knowledge. However, direct in- sence of law enforcement in Serra das Lontras National terviews are prone to non-responses and false answers Park and the buffer zone, and the Park’s more recent desig- (Gavin et al., ; Mgawe et al., ). It may be that the nation as a protected area, could be contributing to the per- proportion of people involved in hunting, and especially ception that hunting was higher and not decreasing there. those actively involved, is underestimated in our results. Law enforcement in the region is irregular and infrequent. The randomized response technique was also a valuable ap- During – the Chico Mendes Institute for proach, as it guaranteed confidentiality to our respondents Biodiversity Conservation recorded  seizures related to and facilitated our investigation of species consumption. hunting (seizure of hunting equipment and animal body The technique works well when behaviours are relatively parts) in Una Biological Reserve and Una Wildlife Refuge widespread and not too sensitive, as was the case for some in  patrol days. However, there were no law enforcement of the species we investigated. However, the method has events during this period in Serra das Lontras National Park drawbacks, including the complex explanations required, or the buffer zone, because of a lack of financial and human especially if the targeted population is illiterate (St. John resources. In addition, after Una Biological Reserve was es- et al., ; Razafimanahaka et al., ). In addition, a tablished, environmental educational activities and tech- large sample size is necessary to produce precise estimates nical support for agriculture were provided for rural (Lensvelt-Mulders et al., ). Our sample size was rela- residents living in areas adjacent to the protected area, tively small, resulting in large standard errors, which which are now designated as the Wildlife Refuge (IBAMA,

FIG. 2 Estimates of species consumption (with standard error bars) in and around protected areas in southern Bahia, Brazil (Fig. ), in the  months prior to interviews conducted using the randomized response technique. The number of respondents who recognized the species and answered the question is indicated in parentheses. Threatened species are indicated with an asterisk (*). ** Estimated proportion ×: estimated % of respondents that have consumed the species (negative values and values above % are due to the probabilistic nature of the method).

). Santos & Blanes () reported a reduction in hunt- against hunting in Brazil, considers the activity to be illegal, ing levels around the Reserve after  years of the environ- without exceptions. As the Chico Mendes Institute is re- mental education programme (initiated in ), which sponsible for addressing administrative infractions in pro- disseminated information about forest conservation and tected areas and buffer zones, hunting is illegal inside the the objectives of the protected area, property rights, environ- study area. According to protected area managers, when a mental laws, and new alternatives for land use. The reasons case of hunting to satisfy hunger is observed during a law why respondents were divided on whether species abun- enforcement event, this information is registered and is con- dances were increasing or decreasing are not clear. It may sidered by the body responsible for taking action with re- be a reflection of local environmental variation, such as in spect to the infraction. Despite only a few respondents presence of roads, human population density and size of for- stating that they hunt occasionally because of a lack of ani- est fragments. Alternatively there may have been differences mal protein, protected area managers need to take this into in how respondents interpreted the question. Further account, and further work should investigate who these oc- research is needed to understand these responses better. casional hunters are, and what proportion of residents hunt With respect to socio-economic factors, we found a rela- to satisfy hunger. tionship between educational level and hunting, with people with primary education hunting more than illiterate people Wild meat consumption (who were generally older) or those with other levels of education. We also found that older respondents consumed col- Our finding that pacas, armadillos and collared peccaries lared peccary more than younger respondents. Without were both the most consumed species and the most pre- more contextual information and validation, however, ferred wild meats confirms the findings of other studies in these relatively weak associations cannot be taken as a the region (Pereira & Schiavetti, ; Flesher & Laufer, basis for policy. Other confounding factors may be respon- ) and nationally (Cullen et al., ). We also found sible for the relationships observed. With respect to hunting that consumption of threatened species was low in the legality, according to the Law of Environmental Crimes, study area (too low for the randomized response technique hunting is illegal in Brazil except ‘in case of necessity to sat- to be able to estimate prevalence robustly). Some species, isfy the hunger of a person or his family, protection of crops particularly monkeys, sloths and porcupines, are not prized and cattle from predatory or destructive animals, and for an- for consumption and are not easy to hunt unless encountered imals identified as harmful’ (provided there has been previ- opportunistically. However, according to records provided by ous authorization; Art.  Federal Law /). However, the Chico Mendes Institute, one of the men fined during a law Decree /, which defines administrative infractions enforcement event in Una Biological Reserve declared that he

TABLE 5 Minimal adequate models (logistic analysis of covariance) TABLE 6 Preferred wild species for consumption among interview showing factors affecting the estimated proportion of consumption respondents (n = ) in three protected areas and a buffer zone of paca, armadillo and collared peccary, using the RRT results. in southern Bahia, Brazil (Fig. ). Threatened species are in bold.

Estimate SE z P(. |z|) Preferred species No. of respondents Model 1 Paca Armadillo 44 (Intercept) −0.7029 0.5071 −1.386 0.166 Paca 42 Protected area (Serra 1.0784 0.6577 1.640 0.101 Opossum 28 das Lontras National Collared peccary 8 Park–Buffer zone) Deer 6 Model 2 Armadillo Coati 4 (Intercept) 1.31438 0.61048 2.153 0.0313* Agouti 1 Duration of residence −0.04416 0.02392 −1.846 0.0648 Collared anteater 1 Model 3 Peccary Southern Bahian masked titi 1 (Intercept) −5.71122 2.72158 −2.098 0.0359* Kinkajou 1 Age 0.08800 0.04547 1.935 0.0530

*P , . for residents, whereas in Una Wildlife Refuge and the buffer zone it is necessary to take an approach to wildlife manage- had hunted yellow-breasted capuchins Sapajus xanthoster- ment that involves local people, such as an expanded educa- nos. The southern Bahian masked titi and the golden-headed tion programme to disseminate information about the lion tamarin offer little meat in return for effort invested and conservation status of threatened species and the rules gov- therefore are probably not killed frequently. Taboos can also erning hunting (Milner-Gulland et al., ; Gandiwa, reduce hunting pressure on threatened and endemic species ). It is also necessary to take account of local needs for   (Colding & Folke, ; Jones et al., ) but are not always protein, even if this involves only a small portion of resi- sufficient; our respondents suggested that when there was a dents. Furthermore, government and conservation institu- need for animal protein and an opportunity for an easy tions must allocate more financial resources to implement hunt (e.g. a maned sloth crossing a road), non-preferred the National Action Plan and protect the threatened species species usually protected by taboos were still hunted. in the study region through effective protected area Although the most preferred, and most consumed, spe- management. cies are relatively common in the study region, hunting can still lead to declines. Cassano et al. () did not find pacas, peccaries or deer (Mazama sp.) in camera traps in the study Acknowledgements region, and suggested that hunting may have caused these species to become uncommon. Species that should occur This study was funded by the Brazilian National Council for in the region were either reported to have disappeared or Scientific and Technological Development (CNPq), the were unknown to our respondents, suggesting that there Centre for Research and Conservation of the Royal may have been hunting-related extirpations in the past. Zoological Society of Antwerp (funded by the Flemish These include, among others, the northern brown howler Ministry for Economy, Science and Innovation) and the monkey, which is probably locally extinct (Flesher & State University of Santa Cruz. We thank the Bahia State Laufer, ; Neves et al., ). Research Support Foundation for providing a PhD scholarship to LCC, and CNPq for providing a PhD sandwich scholarship to LCC to spend time working at the Interdisciplinary Centre for Conservation Science in the Conclusions UK, and a research fellowship to AS. We are grateful to Reducing hunting in protected areas is challenging and the managers of the protected areas; to Camila Cassano, complex and requires an integrated approach. Our findings Natalia Hanazaki, Rômulo Alves, Romari Martinez and suggest that illegal hunting is occurring within and around the anonymous reviewers for their valuable comments; protected areas in the study area but that people mostly want and to the rural communities in the study area for their will- to hunt common species for their own consumption. The ingness to participate in the interviews. relatively low level of consumption of threatened species suggests that reducing hunting pressure on threatened species, one of the goals of the National Action Plan in southern References Bahia (ICMBio, ), is not an impossible task. Management ALVES, R.R.N. () Fauna used in popular medicine in Northeast efforts in Una Biological Reserve and Serra das Lontras Brazil. Journal of Ethnobiology and Ethnomedicine, , https://dx.doi. National Park should prioritize a fair expropriation process org/./---.

ALVES, R.R.N., MENDONÇA, L.E.T., CONFESSOR, M.V.A., VIEIRA, FOERSTER, S., WILKIE, D.S., MORELLI, G.A., DEMMER, J., STARKEY, W.L.S. & LOPEZ, L.C.S. () Hunting strategies used in the M., TELFER,P.etal.() Correlates of bushmeat hunting among semi-arid region of northeastern Brazil. Journal of Ethnobiology and remote rural households in Gabon, Central Africa. Conservation Ethnomedicine, , https://doi.org/./---. Biology, , –. ALVES, R.R.N. & SOUTO, W.M.S. () Ethnozoology in Brazil: FOX, J.A. & TRACY, P.E. () Randomized Response: A Method for current status and perspectives. Journal of Ethnobiology and Sensitive Surveys. Sage Publications, Beverly Hills, USA. Ethnomedicine, , https://doi.org/./---. GANDIWA,E.() Preliminary assessment of illegal hunting by BARBOSA, J.A.A., NOBREGA, V.A. & ALVES, R.R.N. () Hunting communities adjacent to the northern Gonarezhou National Park, practices in the semiarid region of Brazil. Indian Journal of Zimbabwe. Tropical Conservation Science, , –. Traditional Knowledge, , –. GARDNER, C.J. & DAVIES, Z.G. () Rural bushmeat consumption BENNETT, E., EVES, H., ROBINSON,J.&WILKIE,D.() Why is within multiple-use protected areas: qualitative evidence from eating bushmeat a biodiversity crisis? Conservation in Practice, , Southwest Madagascar. Human Ecology, , –. –. GAVIN, M.C., SOLOMON, J.N. & BLANK, S.G. () Measuring and BENNETT, E.L. & ROBINSON, J.G. () Hunting for the Snark. In monitoring illegal use of natural resources. Conservation Biology, , Hunting for Sustainability in Tropical Forests (eds J.G. Robinson & –. E.L. Bennett), pp. –. Columbia University Press, New York, USA. HANAZAKI, N., ALVES,R.&BEGOSSI,A.() Hunting and use of CANALE, G.R., PERES, C.A., GUIDORIZZI, C.E., GATTO, C.A.F. & terrestrial fauna used by Caiçaras from the Atlantic Forest coast KIERULFF, M.C.M. () Pervasive defaunation of forest remnants (Brazil). Journal of Ethnobiology and Ethnomedicine, , http://dx. in a tropical biodiversity hotspot. PLoS ONE, (), e. doi.org/./---. CARVALHO,JR, E.A.R. & MORATO, R.G. () Factors affecting big cat HARRISON, M., BAKER, J., TWINAMATSIKO,M.&MILNER-GULLAND, hunting in Brazilian protected areas. Tropical Conservation Science, E.J. () Profiling unauthorized natural resource users for better , –. targeting of conservation interventions. Conservation Biology, , CARVALHO, E.A.R. & PEZZUTI, J.C.B. () Hunting of jaguars and –. pumas in the Tapajós–Arapiuns Extractive Reserve, Brazilian IBAMA () Plano de Manejo da Reserva Biológica de Una. Http:// Amazonia. Oryx, , –. www.icmbio.gov.br/portal/images/stories/imgs-unidades-coservacao/ CASSANO, C.R., BARLOW,J.&PARDINI,R.() Large mammals in REBIOUna.pdf [accessed  February ]. an agroforestry mosaic in the Brazilian Atlantic Forest. Biotropica, ICMBIO () Executive Summary of the National Action Plan for the , –. Conservation of Central Atlantic Forest Mammals. Http://www. CASTILHO, L.C., MARTINEZ, R.A., GINÉ, G.A.F., RIBEIRO, G.C. & icmbio.gov.br/portal/images/stories/docs-plano-de-acao/ SCHIAVETTI,A.() The thin-spined porcupine, Chaetomys pan-mamiferos-da-mata-atlantica/sumario_mamiferos_mata_ subspinosus (Rodentia: Erethizontidae), within protected areas in atlantica_ingles.pdf [accessed  March ]. the Atlantic Forest, Brazil: local knowledge and threats. Tropical ICMBIO () Planos de Ação Nacional. Http://www.icmbio.gov.br/ Conservation Science, , –. portal/biodiversidade/fauna-brasileira/planos-de-acao-nacional. CHIARELLO,A.G.() Influência da caça ilegal sobre mamíferos e aves html [accessed  March ]. das matas de tabuleiro do norte do estado do Espírito Santo. Boletim JONES, J.P.G., ANDRIAMAROVOLOLONA, M.M. & HOCKLEY,N.() do Museu de Biologia Mello Leitão (Nova Série), /, –. The importance of taboos and social norms to conservation in COLDING,J.&FOLKE,C.() The relations among threatened Madagascar. Conservation Biology, , –. species, their protection, and taboos. Conservation Ecology, , http:// KEANE,A.() Sensitive: Functions for the analysis of sensitive www.consecol.org/vol/iss/art/. survey data collected using specialised questioning techniques. CORLETT, R.T. () The impact of hunting on the mammalian fauna R package version .. of tropical Asian forests. Biotropica, , –. LENSVELT-MULDERS, G., HOX,J.&VAN DER HEIJDEN,P.() CULLEN,JR, L., BODMER, R.E. & VALLADARES PÁDUA,C.() How to improve the efficiency of randomised response designs. Effects of hunting in habitat fragments of the Atlantic forests, Brazil. Quality & Quantity, , –. Biological Conservation, , –. LOIBOOKI, M., HOFER, H., CAMPBELL, K.L.I. & EAST, M.L. () CULLEN,JR, L., BODMER, R.E. & VALLADARES-PADUA,C.() Bushmeat hunting by communities adjacent to the Serengeti Ecological consequences of hunting in Atlantic forest patches, São National Park, Tanzania: the importance of livestock ownership and Paulo, Brazil. Oryx, , –. alternative sources of protein and income. Environmental EL BIZRI, H.R., MORCATTY, T.Q., LIMA, J.J.S. & VALSECCHI,J.() Conservation, , –. The thrill of the chase: uncovering illegal sport hunting in Brazil MELO, R.S., SILVA, O.C., SOUTO, A., ALVES, R.R.N. & SCHIEL,N. through YouTube posts. Ecology and Society, , http://dx.doi.org/ () The role of mammals in local communities living in ./ES--. conservation areas in the Northeast of Brazil: an ethnozoological FA, J.E. & BROWN,D.() Impacts of hunting on mammals in approach. Tropical Conservation Science, , –. African tropical moist forests: a review and synthesis. Mammal MGAWE, P., MULDER, M.B., CARO, T., MARTIN,A.&KIFFNER,C. Review, , –. () Factors affecting bushmeat consumption in the Katavi-Rukwa FERNANDES-FERREIRA, H., MENDONÇA, S.V., ALBANO, C., FERREIRA, ecosystem of Tanzania. Tropical Conservation Science, , –. F.S. & ALVES, R.R.N. () Hunting, use and conservation of birds MILNER-GULLAND, E.J., BENNETT, E.L. & THE SCB  ANNUAL in Northeast Brazil. Biodiversity and Conservation, , –. MEETING WILD MEAT GROUP () Wild meat: the bigger FERNANDES-FERREIRA, H., MENDONÇA, S.V., CRUZ, R.L., picture. Trends in Ecology & Evolution, , –. BORGES-NOJOSA, D.M. & ALVES, R.R.N. () Hunting of MINZENBERG,E.&WALLACE,R.() Amazonian agriculturalists herpetofauna in montane, coastal and dryland areas of northeastern bound by subsistence hunting. Journal of Cultural Geography, , Brazil. Herpetological Conservation and Biology, , –. –. FLESHER, K.M. & LAUFER,J.() Protecting wildlife in a heavily MMA () Sistema Nacional de Unidades de Conservação (SNUC). hunted biodiversity hotspot: a case study from the Atlantic Forest of Http://www.mma.gov.br/port/conama/legiabre.cfm?codlegi= Bahia, Brazil. Tropical Conservation Science, , –. [accessed  November ].

MORCATTY,T.&VALSECCHI,J.() Social, biological, and bushmeat consumption reveals high consumption of protected environmental drivers of the hunting and trade of the endangered species in Madagascar. Oryx, , –. yellow-footed tortoise in the Amazon. Ecology and Society, , REDFORD, K.H. () The empty forest. BioScience, , –. http://dx.doi.org/./ES--. RIBEIRO, G.C., PEREIRA, J.P.R., DOCIO, L., ALARCON, D.T. & MYERS, N., MITTERMEIER, R.A., MITTERMEIER, C.G., FONSECA, G.A. SCHIAVETTI,A.() Zooteráticos utilizados no sul da bahia. In B. & KENT,J.() Biodiversity hotspots for conservation Zooterapia: Os Animais na Medicina Popular Brasileira (eds E.M. priorities. Nature, , –. Costa Neto & R.R.N. Alves), pp. −. NUPPEA, Recife, Brazil. NASCIMENTO, R.A., SCHIAVETTI,A.&MONTAÑO, R.A.M. ()An SANTOS,G.&BLANES,J.() Environmental education as a strategy assessment of illegal capuchin monkey trade in Bahia State, Brazil. for conservation of the remnants of Atlantic forest surrounding Una Neotropical Biology and Conservation, , –. Biological Reserve, Brazil. Dodo, , –. NASI, R., TABER,A.&VAN VLIET,N.() Empty forests, empty SCHIAVETTI,A.,MAGRO,T.C.&SANTOS,M.S.() Implementação stomachs? Bushmeat and livelihoods in the Congo and Amazon das Unidades de Conservação do Corredor Central da Mata Atlântica Basins. International Forestry Review, , –. no estado da Bahia: desafios e limites. Revista Árvore, , –. NEVES, L.G., CLARK, F.E., MELO, F.R., RYLANDS, A.B. & TALEBI,M. SCHROTH, G., FARIA, D., ARAUJO, M., BEDE, L., VAN BAEL, S.A., () Northern brown howler monkey. In Primates in Peril: The CASSANO, C.R. et al. () Conservation in tropical landscape World’s  Most Endangered Primates – (eds C. Schwitzer, mosaics: the case of the cacao landscape of southern Bahia, Brazil. R.A. Mittermeier, A.B. Rylands, L.A. Taylor, F. Chiozza, E. Biodiversity and Conservation, , –. A. Williamson et al.), pp. –. IUCN SSC Primate Specialist SOLOMON, J., JACOBSON, S.K., WALD, K.D. & GAVIN,M.() Group, International Primatological Society, Conservation Estimating illegal resource use at a Ugandan park with the International and Bristol Zoological Society, Arlington, USA. randomized response technique. Human Dimensions of Wildlife, , NIELSEN, M.R. & MEILBY,H.() Determinants of compliance with –. hunting regulations under Joint Forest Management in Tanzania. ST.JOHN, F.A.V., EDWARDS-JONES, G., GIBBONS, J.M. & JONES, J.P.G. South African Journal of Wildlife Research, , –. () Testing novel methods for assessing rule breaking in NUNO, A., BUNNEFELD, N., NAIMAN, L.C. & MILNER-GULLAND, E.J. conservation. Biological Conservation, , –. () A novel approach to assessing the prevalence and drivers of SECRETARIAT OF THE CONVENTION ON BIOLOGICAL DIVERSITY illegal bushmeat hunting in the Serengeti. Conservation Biology, , () Livelihood Alternatives for the Unsustainable Use of –. Bushmeat. Report prepared for the CBD Bushmeat Liaison Group. NUNO,A.&ST JOHN, F.A.V. () How to ask sensitive questions in Technical Series No. , Montreal, Canada. conservation: a review of specialized questioning techniques. WARNER, S.L. () Randomized response: a survey technique for Biological Conservation, , –. eliminating evasive answer bias. Journal of the American Statistical PEREIRA,J.&SCHIAVETTI,A.() Conhecimentos e usos da fauna Association, , –. cinegética pelos caçadores indígenas ‘Tupinambá de Olivença’ WRIGHT, S.J., STONER, K.E., BECKMAN, N., CORLETT, R.T., DIRZO, R., (Bahia). Biota Neotropica, , –. MULLER-LANDAU, H.C. et al. () The plight of large animals in PERES, C.A. () Effects of subsistence hunting on vertebrate tropical forests and the consequences for plant regeneration. community structure in Amazonian forests. Conservation Biology, Biotropica, , –. , –. PERES, C.A. () Synergistic effects of subsistence hunting and habitat fragmentation on Amazonian forest vertebrates. Biographical sketches Conservation Biology, , –. PERES, C.A., EMILIO, T., SCHIETTI, J., DESMOULIÈRE, S.J.M. & LEVI,T. L UCIANA C ASTILHO is interested in social-ecological research, inves- () Dispersal limitation induces long-term biomass collapse in tigating the use of natural resources, and particularly wild fauna, in pro- overhunted Amazonian forests. PNAS, , –. tected areas, and its implications for wildlife conservation and the PERES, C.A. & NASCIMENTO, H.S. () Impact of game hunting by welfare of local people. K RISTEL D E V LEESCHOUWER is a primate the Kayapó of south-eastern Amazonia: implications for wildlife ecologist studying factors affecting the persistence of golden-headed conservation in tropical forest indigenous reserves. Biodiversity and lion tamarin populations in degraded and anthropized landscapes in Conservation, , –. south Bahia, and working towards science-based conservation planning PERES, C.A. & PALACIOS,E.() Basin-wide effects of game harvest for the species. E. J. M ILNER-GULLAND leads the Interdisciplinary on vertebrate population densities in Amazonian forests: Centre for Conservation Science at Oxford University (http://www. implications for animal-mediated seed dispersal. Biotropica, , iccs.org.uk) and has a particular interest in understanding the motiva- –. tions of local resource users. A LEXANDRE S CHIAVETTI’s research in- RAZAFIMANAHAKA, J.H., JENKINS, R.K.B., ANDRIAFIDISON, D., terests include planning and management of protected areas, especially RANDRIANANDRIANINA, F., RAKOTOMBOAVONJY, V., KEANE,A.& for solving conflicts between users of natural resources and biodiversity JONES, J.P.G. () Novel approach for quantifying illegal conservation.