Samples were soaked in a neutral mixture of dilute hydrogen peroxide and Calgon for one to six hours. Chalky samples were manually broken up after an initial period of soaking to assist complete disaggregation. All samples below the upper Miocene level were freeze dried before soaking. Disaggregated material was wet sieved through a 63-µm screen, and the >63-µm fraction was dried in an oven at <24°C. The material was then dry sieved through a 150-µm screen and the fractions were archived separately.
The >150-µm fraction was examined thoroughly in order to include rare species. At least two trays of sediment were inspected. Relative abundance estimates are based on the >150-µm fraction. Three categories of abundance were employed: rare (1%-3%), few (4%-15%), and common (16%-30%). Preservation of foraminifers was classed as either good (>90% of specimens unbroken; few signs of dissolution), moderate (30%-90% of specimens unbroken or dissolved), or poor (>90% of specimens broken or dissolved). No samples were examined that meet the ODP criteria for "excellent" preservation.
Generally, three samples per core were examined in detail. Those samples were usually drawn from section 2, section 5, and the core catcher. Datum levels are estimated by interpolation between these sample depths. The ages of samples are determined by interpolating between planktonic foraminifer datums that have reliable age estimates. Reliability is judged by the relative invariability of a datum from site to site at well-dated sites with good preservation and recovery.