Abstract
Track tubes are a noninvasive, efficient method to monitor populations of small mammals that can be implemented on a large landscape scale and are a cost-effective approach for certain sampling situations. As with all field sampling tools, modifications are made depending on research objectives, habitat being sampled, and target species. We conducted two experiments with the objective to increase efficiency and decrease labor while retaining high detection probabilities as part of an Annual Multi-Agency Regional Southeastern Beach Mouse Peromyscus polionotus niveiventris Habitat Occupancy Survey. We conducted studies along a contiguous 72-km coastline of the Cape Canaveral Barrier Island Complex in east-central Florida, USA, that includes the Kennedy Space Center, Cape Canaveral Air Force Station, and Canaveral National Seashore. We conducted the experiments to address several issues that had regularly compromised our survey data. One experiment assessed the ideal length of deployment times for track tubes in various habitat types with multiple species where southeastern beach mice were previously detected. Another experiment, determined how to restrict certain meso-mammal species (raccoons Procyon lotor, and eastern spotted skunks Spilogale putorius) from knocking over the tubes or reaching into the tubes, thus reducing disturbance that results in missing data; or from entering the tubes (cotton rats Sigmodon hispidus), thereby obscuring any potential or actual footprints of the southeastern beach mouse or other targeted species. We observed an increase in the detection rate of southeastern beach mice in track tubes with increased nights of deployment (likelihood ratio test χ2 = 18.71, df = 3, P < 0.001) with the greatest increase between 3 nights and 6 nights and apparent leveling off between nights 9 and 12 and a large decrease in the detection rate of southeastern beach mice in track tubes with 0.5-in. (1.3-cm) excluders compared with the other-size excluders (likelihood ratio test χ2 = 167.89, df = 5, P < 0.001). We also found that 1-in. (2.5-cm) excluders prevented access by meso-mammals or entry by cotton rats and did not adversely affect the detection of either beach mice or cotton mice Peromyscus gossypinus, but did reduce disturbance and resulted in fewer missing data. These statistically significant process improvements have application for others involved with small mammal monitoring and species management.
LIKELIHOOD RATIO TEST FOR MEAN VECTOR WITH SPECIFIED DIRECTION BASED ON CONDITIONAL DISTRIBUTION