Uso de marcadores moleculares na identificação de espécimes atropelados na Amazônia: uma contribuição para a ecologia de transporte e conservação da biodiversidade

Abstract

Transportation Ecology has demonstrated considerable significance in elucidating how transport infrastructures impact biodiversity. Accordingly, identifying roadkill hotspots and the most affected species within specific areas is crucial for implementing effective mitigation strategies. Traditionally, species identification from highway roadkill is based on morphological characters; nevertheless, the condition of carcasses and the high species diversity can complicate this process. In this context, molecular identification via DNA barcoding can assist in taxon determination. However, this approach also faces limitations, notably DNA degradation in carcass-derived samples and the absence of reference sequences for many species. Against this backdrop, the present study aimed to evaluate DNA extraction from roadkill samples by comparing three protocols, assess DNA degradation levels, and amplify the COI, 16S rRNA, and 12S rRNA barcode regions to determine their discriminatory power across taxa. The specimens analyzed belong to the Reconecta Project collection, surveyed along a stretch of the BR-174 highway in the Amazon region. Initially, DNA extraction was performed using a saline-buffer protocol, with universal COI primers applied for amplification. For samples that failed to amplify, two alternative extraction methods and additional barcoding markers were tested. Results revealed highly degraded DNA likely impeding successful amplification of the ~670bp COI fragment. In contrast, the barcode (~550bp) and mini-barcode (~98bp) regions of the 16S rRNA and 12S rRNA genes, respectively, were efficiently amplified across all extraction protocols. Among the markers evaluated, the 16S rRNA region yielded the highest number of high-quality sequences, even from extensively degraded samples. Both COI and 16S rRNA data enabled discrimination of most mammalian species, although resolution for reptilian taxa was hampered by insufficient reference sequences in public databases. Overall, the saline-buffer extraction method proved to be adequate for most roadkill samples. However, amplification success varied by marker type, with the 16S rRNA barcode emerging as the most robust choice, producing consistent positive amplification in the majority of samples. The comparative analysis of morphological versus molecular identification underscored the complementary utility of both approaches in Transportation Ecology research.