Fatores da paisagem que influenciam a abundância e a detectabilidade do tamanduá-bandeira (Myrmecophaga tridactyla) em seis unidades de conservação do Brasil

Abstract

Abundance and density estimates are needed to better understand a given population and to generate conservation strategies. The giant anteater (Myrmecophaga tridactyla) is the largest existing myrmecophagous, has a wide distribution in the Neotropics, and is vulnerable to extinction due to an estimated loss of at least 30% of its population in the last three decades. The species is naturally rare and there are few studies that estimate its abundance and population density, and this topic is considered one of the knowledge gaps about the species. The goals of this study were (1) to estimate the abundance of giant anteaters on a finer scale (ie, sampling point) and to assess the landscape factors that influence this abundance and the probability of individual detection at the sampling points using N-mixture models; (2) estimate the abundance and population density of giant anteaters for each conservation unit using spatially explicit capture-recapture (SECR) models. We carried out the study with camera trap data for six protected areas (PAs) in Brazil. We analyzed the influence of variables from the protected area scale and from the sampling point scale, totaling 16 variables, on the average abundance and detection of giant anteaters. The average abundance (λ) of giant anteaters in the sampling points was negatively influenced by the index of proximity to savanna formation (proxsa), probably due to the low heterogeneity of habitats that surround the studied PAs. The average abundance in the sampling points ranged from 0.63 ind./sampling point (CI-95% = 0.35 – 1.16) for the APA Cochá-Gibão to 1.85 ind./sampling point (CI-95%) = 1.22-2.82) for the Rio Preto State Park. The probability of detecting individuals in traps installed on trails was greater than outside them, and the greater the distance from the sampling point to an unpaved road, the greater the probability of detecting an individual of giant anteater. The positive effect of the distance to unpaved roads is probably related to a greater occupation of people and domestic animals near them, as these roads are located mainly in the area surrounding the sampled PAs. Regarding the estimates generated by the SECR models, the study area with the highest density was the Sempre Vivas National Park, with an estimate of 0.031 ind./km² and the lowest was the Serra do Cabral State Park, with only 0. 0115 ind./km². But, in general, the values of the density estimates are close and the confidence intervals overlap. These are the first estimates of abundance and population density for almost all of the PAs analyzed, contributing to fill one of the knowledge gaps about the species, which is the lack of population estimates. This was the first work to use N-mixture models to estimate the population abundance of giant anteaters and this could be a good method to be used in a future unified protocol that allows the comparison of population estimates between different areas and allows monitoring giant anteater populations over time, which will be essential for their future conservation.