Community occupancy of herpetofauna in roadside ditches in a managed pine landscape

Aquatic habitat types embedded in managed forest can contribute ecological services and suitable conditions for numerous species despite departing from historical baselines. Although prevalent in the Atlantic Coastal Plain, roadside ditches adjacent to pine plantations have had few surveys of vertebrate diversity and little is known about local and landscape level effects of silvicultural activities on species assemblages. Our goal was to investigate richness and occupancy of amphibians and reptiles in an intensively managed forest landscape with a history of silvicultural alterations of aquatic habitat types, including ditching, draining, and periodic maintenance of ditches for optimizing loblolly pine (Pinus taeda) growth. We selected 15 roadside ditches that were maintained 3-17 years earlier and were embedded in a matrix of pine plantations in eastern North Carolina, USA. We conducted repeated, visual encounter surveys of roadside ditches during 2012-2013 and detected 25 species of amphibians and reptiles 447 times including five species of statewide conservation concern. We used a Bayesian, hierarchical, community occupancy model with group-specific hyper-parameters to estimate occupancy probabilities and species richness while accounting for imperfect detection. With our model, we examined effects of time since ditch maintenance, amount of nearby mature forest, and amount of adjacent wetlands on species richness and occupancy. We predicted species would respond positively to an intermediate period of time since ditch maintenance and positively to landscape covariates related to mature forest and wetlands. Our model estimated species richness of anurans from 0.9 to 7.4 species/ditch segment, snakes and lizards from 0.8 to 9.0 species/ditch segment, and turtles from 0.9 to 5.1 species/ditch segment. Contrary to our predictions, occupancy of herpetofauna showed no evidence of a relationship to time since ditch maintenance, and landscape metrics only influenced occupancy of one turtle species. Detection probabilities were influenced by season for several anurans, snakes, and lizards, but not turtles. Our results indicate that aquatic habitat types embedded in managed forests that were hydrologically and structurally altered for silviculture can support local occupancy of a relatively diverse herpetofaunal community and that temporal proximity to ditch maintenance had little effect on occupancy or richness.