Shifting cultivation and hunting across the savanna-forest mosaic in the Gran Sabana, Venezuela: facing changes

Background Human encroachment and overexploitation of natural resources in the Neotropics is constantly increasing. Indigenous communities all across the Amazon, are trapped between a population rise and a hot debate about the sustainability of hunting rates. The Garden Hunting hypothesis states that shifting cultivation schemes (conucos) used by Amazon indigenous communities may generate favorable conditions, increasing abundance of small and medium wildlife species close to the ‘gardens’ providing game for indigenous hunters. Methods Here, we combined camera trap surveys and spatially explicit interview dataset on Pemón indigenous hunting scope and occurrence in a mosaic of savanna and forest in the Gran Sabana, Venezuela to evaluate to what extent the wildlife resource use corresponds to Garden Hunting hypothesis. We applied the Royle–Nichols model and binomial regression in order to: (1) assess whether abundance of small and medium wildlife species is higher close to conucos and (2) evaluate whether hunters select hunting localities based on accessibility to wildlife resources (closeness to conuco) more than wildlife abundance. Results We find mixed evidence supporting the Garden Hunting hypothesis predictions. Abundance of small and medium species was high close to conucos but the pattern was not statistically significant for most of them. Pemón seem to hunt in locations dominated by forest, where species abundance was predicted to be higher, than in close vicinity to conucos. Hunting scope was focused on the most abundant species located close to the conuco (Cuniculus paca), but also in less abundant and unavailable species (Crax alector, Tapirus terrestris and Odocoileus virginianus). Conclusions Our research provided the first attempt of a systematic sampling survey in the Gran Sabana, generating a quantitative dataset that not only describes the current pattern of wildlife abundance, but sets the base-line to monitor temporal and spatial change in this region of highland Amazon. We discuss the applicability of the estimates generated as a baseline as well as, environmental challenges imposed by economic, social and cultural changes such as mining encroachment for wildlife management.

212 the camera trap survey (see next section) looking for evidence of hunters, hunters with firearms 213 or dogs. In each sampling unit where the camera trap was installed, we asked whether they hunt 214 (1) or not (0) to obtain spatial distribution of hunting occurrence in the study area. To accurately 215 identify animals hunted, and avoid misinterpretation with animals' local names, we showed 216 pictures and illustrations of wildlife (Linares, 1998) to the participants. To identify which species are important game species for Pemón people, we used two 219 criteria: the frequency a given species is reported as target game, and how frequently it is 220 mentioned as preferred game. For that we calculate two indexes for each species, importance of 221 hunting (Hv) and hunting preference (Pv) (Carvalho et al., 2015). Both indexes correct the bias 222 introduced by sampling size in the species citation rate, by multiplying the number of informants 223 giving information on each species. Hv is defined as: The hunting preference index (Pv), measure the frequency that each species is cited as the 230 first option for hunting among others, and is defined by: where p is the number of times a species is cited as the first option (among the three most 233 preferred hunted species), n the total number of citations for all prey species, and N the number 234 of interviewees. In this case, zero values (i.e. no preference) were excluded. During fieldwork we marked with a GPS the location of active and recently abandoned 305 conucos (n=25) identified in situ, and hunting sites (n=32) reported by interviewees and 306 confirmed by the local guides ( Fig. 1). Distance from each camera to the nearest conuco was 307 calculated using the GPS coordinates from cameras and conucos. This variable had an 308 asymmetric distribution with a mean value of 1.58 km and a range from 0 to 8 km. We also 309 recorded which cameras were located adjacent or near reported hunting sites (binomial variable 310 hunting, FALSE n = 23, TRUE n = 34).

312
After visual inspection of the distribution of tree cover, hunting occurrence, and distance to 313 conuco, we decided to discard three cameras with extreme distance values. We also discarded 314 four cameras that were active for less than seven days. Thus all following analyses count on data 315 from 54 cameras within 5 km of the nearest conuco, with more than seven days of activity ( Fig.   316 2). To evaluate the prediction of Garden Hunting Hypothesis we followed a three-step 321 approach using occupancy model, Chi-square test and logistic regression.

323
To test the prediction whether abundance of small and medium wildlife species is 324 higher close to conucos, we need a measure of influence of conucos while controlling for the 325 influence of habitat on species abundance, and the spatial and temporal heterogeneity in

Frequency of detection of mammals and birds
In general, and despite few outliers, abundance predictions from the mixture models were 496 higher for most species in sites where the Pemón reported hunting activity (Fig. 5). This was true 497 for species with different values of hunting preference (Hv) and for species not mentioned as 498 important prey for Pemón, including carnivores (with the exception of crab-eating fox; Fig. 5). Hunting occurrence in the study area was detected up to 5 km distance from conucos, both in 503 the savanna and forest (Fig. 2, Supplementary Material 1). Sampling units with reported hunting 504 activity were mostly located at 2.5 km from nearest conucos with tree cover > 40% (Fig. 2). Tree 505 cover (p = 0.006) and distance to rivers (p = 0.070) had a positive significant effect on the 506 hunting occurrence, but the effect of distance to conuco (p = 0.202) was negative and not 507 significant (Table 4).

509
Forest was the preferred hunting area for the majority of interviewees (72%), followed by 510 savanna (31%), and mixed forest -savanna areas (34%). This pattern was similar across 511 communities (χ 2 = 7.67; degree freedom = 6; p ≤ 0.263; Table 5). The majority of interviewees 512 hunt during the rainy season (68%), between May and August, while only 21% interviewees hunt 513 all year round, and 11% had not a preferred season to hunt. This pattern was similar across 514 communities (χ 2 = 9.89; degree freedom = 6; p ≤ 0.129) (Table 5). The vegetation type had the most significant role explaining the abundance pattern of both 538 herbivores and carnivorous species in the study area (Table 3; Stachowicz et al., 2020). Most 539 species modeled, except the crab-eating fox, were more abundant in areas with higher cover 540 trees, which may correspond with forest and shrublands ( As expected, the medium and small species, with fast growing rates like the lowland paca 553 and the red-rumped agouti, had the highest frequency of detection (Table 1). The gray brocket, 554 although less frequent, was yet a prevalent species in the area, which contrast with the almost 555 absence of other deer species, the white-tailed deer (Table 1). Formerly widely distributed and 556 abundant, the white-tailed deer was only detected four times across the six survey months ( Evidence supporting the Garden Hunting prediction about higher abundance of small and 570 medium wildlife species close to conucos was not conclusive: Although most of the herbivorous 571 species modeled seem to have higher abundance close to conucos (Fig. 4) this effect was not 572 significant, and the only two species significantly attracted by conucos were tayra and lowland 573 paca. This pattern seems to agree with previous results on which tayra does not show a marked 574 preference for any type of habitat, while lowland paca showed significant preference to 575 shrublands or intermediate habitat, which likely correspond to conuco (Stachowicz, et al., 2020).

577
Among the species not attracted by conucos, black curassow was the only showing a 578 significant effect (Fig. 4), which contrasts with previous evidence in lowland Amazon and Piaroa 579 communities where cracids were observed within conucos in high abundance (Zent, 1997). 580 These difference might be due to different relationship between indigenous communities and 581 this bird species (with Piaroa using this species as a pet, while for Pemon is a game species), or 582 different habitat preferences of the species between lowland and highland Amazon. For other 583 species of curassow, the Endangered red-billed curassow (Crax blumenbachii), in Brazil was 584 more persistent in forest patches faraway from settlements, with hunting pressure potentially 585 exerting more influence on population persistence than habitat quality (Rios et al., 2020). Again, 586 more detailed population studies are required to better describe population status of this and 587 other endangered species in Gran Sabana, as well as improve our understanding of landscape 588 transformation and human activities in their population dynamics (BirdLife International, 2016).

590
Pemón's hunting occurrence and practice 591 We did not find support for the predictions of higher occurrence of hunting close to conucos 592 (Table 4). Pemón seem to hunt in locations dominated by forest (Table 4 and 5), where species 593 abundance was predicted to be higher (Fig. 5), than in close vicinity of conucos. To our 594 knowledge, there are no studies describing the size of Pemón hunting territories. Here, we found 595 that hunting activity was mostly focused on a radius of 2.5 km from conuco, but we have limited 596 data to test long range hunting (>5 km) (Fig. 2). Evidence from other Pemón community, 597 Tuauken located at ~ 30 km from study area, describes three types of hunting trips (Urbina,