While wildlife management has been perceived as key missing tool for sustainable use of Neotropical resources56,57, few studies have been published on the subject to date3,58. In this study, we have shown that: (1) the total number of harvested snakes, the number of hunters, and the hunting season duration decreased since the implementation of Programa Curiyú in 2002; (2) capture per unit of effort has increased; (3) the number of captures in the most productive month (July) has decreased over time, while the proportion of captures occurring in this month (relative to the total number of anacondas harvested each year) has increased; (4) mean skin length from captured anacondas has not varied significantly, even though more females are being hunted than males. The management framework has therefore proven its usefulness and applicability in a low-cost context, with a hard-to-sample species, leading to its sustainable exploitation.
Most snakes have secretive habits and to estimate their population parameters requires expensive and time-consuming field work, therefore, data obtained from management programs represent a viable source of inference allowing us to identify safe levels of exploitation36,37,49. Over time, decrease in total number of harvested individuals, changes in length and sex ratio36,37,47, or increase in hunting effort, are appropriate indicators of overexploitation. A study analyzing the management of reticulated pythons (Malayopython reticulatus) in Indonesia argues that, for large constricting snakes, the minimum sample required for statistically viable detection of population changes would be 551 individuals, an unlikely feat for most field studies. Additionally, the authors suggest that an annual decline of 5% over approximately 5,000 individuals or declines greater than 10% over smaller samples of about 1,000 individuals, may indicate overexploitation37. In this study—by means of a sample of 54,950 individuals—we conclude that the yellow anaconda harvest is sustainable, since there were no significant declines on analysed parameters.
Previous studies suggest that yellow anacondas are resistant to intense exploitation29,30, as are pythons (Malayopython reticulatus) in Indonesia37 and tegu lizards (Salvator spp.) in Argentina59. Several aspects of their biology contribute to the ability of some reptile species to resist such harvesting pressures, including60: 1) high reproductive rate, 2) wide geographic distribution, 3) strong population inter-connectivity, 4) existence of areas with no hunting, 5) exploitation of disturbed habitats, and 6) low impact hunting techniques (visual search and manual capture). Individual yellow anacondas grow rapidly, have early maturation—attained at the age of 2.5 years—and a relatively high reproductive rate (an average of 24, offspring, though up to 40 has been recorded30), allowing rapid recruitment of new individuals in a short period of time. In addition, juvenile mortality is lower in anacondas than in most other snake species since they are aquatic, very cryptic and aggressive, and are born with a well-developed SVL (40-59 cm) and weight (61-135 g)30. The species range occupies some 42 million hectares, extending from latitudes of 15°S to 30°S30,45,61–63, and encompassing four countries, which makes it unlikely that harvesting in La Estrella could cause total extinction of the species, even if unsustainable. Despite Argentine populations of yellow anacondas show significant genetic structure for mitochondrial genes, males are the principal agents of gene dispersal64. By inhabiting large river courses (La Plata basin is the fifth largest on Earth), these animals are part of an extensive and continuous ecological system that connects the Brazilian Pantanal wetlands to the Argentine Chaco; as a result, their high motility results in populations being genetically structuring as part of a metapopulation30,65.
Additionally, the environmental history of La Estrella marsh, and the conditions of the habitat favor the safe hunting of the anacondas in this part of northeastern Argentina. Covering some 3,000 km², this area represents only 0.70% of the species distribution. Access for hunters to different parts of the marsh is affected by a number of factors, including variable water depth, dense floating vegetation, muddy soils, and fallen trunks40. Therefore, when searching for snakes the hunters paddle their canoes across the marsh using a series of regular navigational routes that are concentrated around the villages. Consequently, large portions of the habitat used by anacondas at La Estrella are not accessed, and this likely create a source-sink dynamic30. Finally, the rudimentary technique used to hunt snakes (visual search and manual capture) is unlikely to be robust enough to have a significant impact on the local anaconda population.
When the capture effort is constant annual harvest reduction is an appropriate parameter to show overharvesting is occurring in a population. Despite an observed decrease in annual captures, this is related to the reduction in hunting effort, and not to a decline in the availability of anacondas in the marsh. There was a significant reduction in the number of accredited hunters, as well in the hunting season duration, between 2002-2016 (Fig. 3), which contrasts with the trending increase in CPUE (Fig. 4). Within the Programa Curiyú there are two types of hunters: the ones that hunt for extra financial income, and the ones whose sole income comes from hunting. The reduction in the hunting effort was mainly due to the abandonment of the activity by people who were not exclusively dedicated to anaconda hunting46. Hunters are a key component in the production chain and their compromise to hunt can be decisive for the success of the management program. Furthermore, understanding that is critical for the correct interpretation of our data, both to prevent inappropriate or unnecessary management interventions and to prevent population declines of the target species. In the current study, the increase in CPUE when hunting effort declined suggests strongly that the abundance of anacondas remains constant in La Estrella.
For snakes in general, long-term analysis of the sex ratio of a population is key to understanding the effects of harvesting programs36. Females represent a critical limiting factor for the persistence of populations, therefore, to ensure long-term availability of a species, hunting programs focused on non-reproductive males would, in overall, be more appropriate than those focused on reproductive females37. So far, during the Programa Curiyú, more females than males have been hunted. However, snakes of the genus Eunectes copulate in aggregations where one female is inseminated by several males, thus, a reduction in the number of males has a positive effect in the number of infertile eggs in females11. Moreover, males have higher dispersion compared to females and as a result are more susceptible to natural predation11. Consequently, male harvest may have unexpected effects and unintended consequences on the reproductive biology of the species.
During the historical period of unrestricted anaconda hunting, before the creation of the Programa Curiyú, specimens above 135 cm of SVL were harvested29,49, indicating that large numbers of young non-reproductive males and females were captured. With the establishment of the Programa Curiyú, an increase in the minimum skin size to 230 cm of SVL was implemented, corresponding to a live specimen of about 200 cm of SVL. The increase in the minimum size restriction resulted in a reduction of production volume by 50-60%, but ensured the protection of a large number of immature and young females49. At first, this rule may seem detrimental to the profitability of hunters and the program, however, larger skins are wider and have bigger scales, which represents a greater monetary value of the resource in the market49.
Although a tend was observed in the analysis of the size of giant skins for a reduction in size over time, we must emphasize that this is an artifact of the data from 2014 and 2015, when the largest hunted individuals were not as big as in the preceding years (see Supplementary Information), with 2016 measurements returning to previous mean values. This may be an effect of the severe drought in 2013. Giant individuals are more likely to die from overheating because they cannot find suitable places to maintain adequate body temperatures, and may face mobility issues when attempting to leave drying mud pools and head towards shading vegetation11. Thus, we argue that there is no tendency to reduction in giant skin sizes, as this is likely a byproduct of the impact of drought on the giant-sized individuals in the local anaconda population.
Environmental factors are thus key determinators of the survival of this species, as well as for the successful development of Programa Curiyú35,62. As a consequence of the 2013 drought, the hunting was prohibited during this particular year, with a subsequent decline in Programa Curiyú productivity, the abandonment of this activity by hunters who sought alternative sources of employment, and a reduction of income of those hunters who remained accredited to the program35,46,66.
The Programa Curiyú had solid empirical evidence that cold fronts improve success when hunting for yellow anaconda. However, there was no scientific evidence showing the relationship between air temperature and captures until now. Thermoregulation is fundamental for survival of ectotherms67, and reptiles can use behavioral and physiological mechanisms68,69, body mass70, and habitat selection71 to regulate body temperature. Yellow anacondas show different body temperatures in terrestrial and aquatic environments72, with body temperature in terrestrial habitats reaching 3-4 °C higher than in water72. Thus, under natural conditions, yellow anacondas can improve heat absorption by getting out of the water and basking on vegetation, but, when doing so, they also become more exposed to being harvested. Knowing that lower temperatures favor the hunt, new strategies can be developed, such as concentrating efforts during colder periods.
Overall, our data corroborate the hypothesis that exploitation of yellow anaconda in northeastern Argentina is sustainable. We conclude, therefore, that the parameters chosen by the Programa Curiyú for hunting restriction are successful in ensuring the demographic viability of the species. The regulation and supervision of management practices by competent institutions and authorities is capable of bringing not only monetary and social benefits to traditional communities, but also to the exploited species by monitoring and conserving its population viability and habitat. Moreover, wildlife management plans represent a great opportunity for scientific exploration of many biological aspects of otherwise difficult-to-survey species, such as the yellow anaconda. Finally, the program exemplifies the use of powerful and cheap tools for sustainable wildlife development, which can be developed for other anaconda populations, as well as for other species and regions.