Evidence of extensive home range sharing among mother-daughter bobcat pairs in the wildland-urban interface of the Tucson Mountains

Urbanization impacts the structure and viability of wildlife populations. Some habitat generalists, such as bobcats (Lynx rufus), maintain populations at the intersection of wild and urban habitats (wildland urban interface, WUI), but impacts of urbanization on bobcat social structure are not well understood. Although commonly thought to establish exclusive home ranges among females, instances of mother-daughter home range sharing have been documented. We combined GPS localities with genomic relatedness inferences from double-digest restriction site associated DNA sequencing (ddRADseq) to investigate mother-daughter home range sharing in bobcats (n = 38) at the WUI in the Tucson Mountains, Arizona, USA. We found the highest relatedness among females, which showed stronger isolation by distance than males and the population as a whole. Using mother-daughter relationships inferred from pedigree reconstruction, we found extensive mother-daughter home range sharing, compared to other females. Every mother identified as having at least one daughter, shared home ranges with one daughter, while other confirmed daughters established adjacent home ranges. Our results provide substantial support for the mother-daughter home range sharing hypothesis, as well as evidence of spatiotemporal overlap between mothers and daughters, adding to the body of research complicating the solitary felid paradigm. These results additionally challenge the notion of home range partitioning by prior rights land tenure, suggesting a role of matrilineal land tenure in home range establishment of daughters. Habitat fragmentation due to human population growth and urbanization thus has the potential to alter landscape genetic structure and social dynamics in bobcats.

Methods
This dataset contains double digest restriction-site associated DNA sequencing data from 38 bobcats (Lynx rufus) sampled between 2020 and 2023 at the wildland-urban interface of the Tucson Mountains, Arizona, USA. Initial processing of raw sequencing data began with removal of PCR duplicates with the Stacks v. 2.60 clone_filter module, followed by trimming an additional five bp from the 5’ ends of the reverse reads using Trimmomatic v. 0.39 (to expose the EcoRI cut site at the front of the read). Finally, we demultiplexed, quality filtered with default settings, and filtered for missing restriction enzymes sites (for enzymes NlaIII and EcoRI-HF) and adapter contamination using the Stacks process_radtags module (using options -c -q -r). Each individual has four resulting fastq (.fq) files in the dataset, including paired reads (files ending in ".1.fq.gz"/".2.fq.gz") and reads remaining after the other read from a read pair was discarded (files ending in ".rem.1.fq.gz"/".rem.2.fq.gz"). The population map file used to process the fastq files with the Stacks denovo_map module is also included. This dataset includes fastq files for 6 individuals later filtered out due to missing data (these are denoted with "yes" in the "Filtered_out" column of the csv file). The final SNP set is included here as a .vcf file, along with a .csv file containing relevant sample metadata and the life history data used for pedigree reconstruction with Sequoia.