PT - JOURNAL ARTICLE AU - Stephen E. Rice AU - Rulon W. Clark TI - Factors affecting intra-island genetic connectivity and diversity of an abundant and widespread island endemic, the San Clemente Island night lizard AID - 10.1101/060038 DP - 2016 Jan 01 TA - bioRxiv PG - 060038 4099 - http://biorxiv.org/content/early/2016/06/20/060038.short 4100 - http://biorxiv.org/content/early/2016/06/20/060038.full AB - Habitat fragmentation and loss are fundamental conservation concerns. Because protected areas are often “habitat islands” isolated by human development, studies of physical islands which document genetic patterns associated with natural versus anthropogenic fragmentation may provide insight into the management of anthropogenic habitat islands. The San Clemente Island night lizard, Xantusia riversiana reticulata, is a highly abundant endemic to two California Channel Islands (San Clemente and Santa Barbara Islands), each with a history of anthropogenic disturbance, and was recently delisted from the Endangered Species Act. We genotyped 917 individuals at 23 microsatellite loci to estimate genetic divergence between collection sites and identify natural and anthropogenic landscape features correlated with genetic divergence for X. r. reticulata throughout its range. Genepool delineation resulted in each well-sampled collection site acting as a distinct population, pairwise Fst values were small but significant, and Structure detected 2 populations on San Clemente Island and 4 populations on Santa Barbara Island. Active sand dunes on San Clemente Island acted as a barrier whereas cholla phase maritime desert scrub, secondary roadways, and canyons > 500ft in length acted as resistant surfaces. Resistant surfaces on Santa Barbara Island were woolly seablite, crystalline iceplant, barren ground, and herbaceous vegetation. Conductance models included California boxthorn and prickly pear cactus on both islands with additional minor features unique to each island. We recommend the application of conductance and resistance models within habitat islands to better inform management for species persistence when the intervening habitat between reserves cannot be improved.