Mammal metabarcoding of western lowland gorillas (Gorilla gorilla gorilla) fecal samples (N=46) from the Dzanga–Sangha Protected Areas, Central African Republic

Since duikers are largely folivorous and mammal DNA can be detected from leaf swabs, we ran a control experiment applying a mammal-generic metabarcoding approach to feces collected from strictly herbivorous and mostly folivorous western lowland gorillas (Gorilla gorilla gorilla; N=46). Gorilla feces yielded no non-host mammal detections, suggesting most non-duiker mammal DNA found in duiker feces represents meat consumption rather than surface eDNA that passed through the gut. Feces was collected from the Dzanga–Sangha Protected Areas (DSPA), Central African Republic (CAR). DSPA lies in southwestern CAR within the Sangha Trinational (with Lobéké, Cameroon, and Nouabalé-Ndoki, the Republic of the Congo), encompassing lowland evergreen and swamp forests with bai clearings. Gorillas are habituated for tourism and research and followed daily, enabling health monitoring and rapid detection of wildlife carcasses in their territories. Gorilla fecal sampling was performed at Mondika, where feces were collected from habituated groups Kingo, Buka, and Metetele; at Goualougo, from the Loya group. Fecal samples were collected using the clean end of a stick to scoop a pea-sized amount of feces into a cryo tube containing NAP buffer. Samples originated from freshly deposited feces collected directly after defecation or beneath recently vacated nests and are therefore unambiguously attributable to known gorillas, making species verification unnecessary. Samples were stored at ambient temperature until their transport to Germany where they were transferred to -20 °C. For the extraction of gorilla feces (N = 46), we used the GeneMatrix Stool DNA Purification Kit (Roboklon, Berlin, Germany) following the manufacturer´s protocol, with the following modifications. Lentil-sized fecal aliquots were transferred into bead tubes with sterile metal spoons, and homogenized using a Bead Ruptor Elite II (Biolabproducts, Hamburg, Germany) for three 20 second cycles at 4m/s. The final elution volume was 100 µL. For all sample types, each extraction batch included a nuclease-free water control. These extraction blanks were carried through all subsequent analyses. DNA was quantified with a Qubit 3 Fluorometer, using either the 1x dsDNA High Sensitivity (HS) or Broad Range (BR) assay kits (Thermo Fisher Scientific, Waltham, Massachusetts, USA). After extraction, work surfaces were decontaminated with 10 % bleach, Terralin®, and UV light for 15 minutes (min). Prior to feces extraction, an additional decontamination step with 10% bleach was carried out, due to the high sensitivity to contamination in mammal metabarcoding analyses. Metabarcoding of gorilla feces was performed using the TagSteady library-preparation workflow to minimize tag-jumping and cross-sample contamination. Tagged mammal-generic 16S mam1/mam2 primers were used together with a single human-specific blocking primer. PCRs were run in triplicate, and amplicons were pooled by plate with positive and negative controls. After end-repair, libraries were ligated to dual Illumina Y-adapters (P5–P7) and sequenced on an Illumina iSeq 100. Negative controls were sequenced alongside samples to monitor potential contamination. Here, we upload the paired fastq files for the three pools from the gorilla diet metabarcoding experiments. Details about the tag combinations are included in the respective csv files, that allow assignment of individual sequences to their respective samples.