The study material from IODP Site 1407 (Hole A, Core 22, 208.10 to 211.85 CCSF; Norris et al. 2014) is part of ongoing Palaeogene climate research. The studied sediments consist of 50–65% CaCO₃, with the remaining 50–35% being mostly opal-CT, clinoptilolite (zeolite) and clays. The samples initially underwent three cycles of wet sieving and oven drying, yet the microfossil content remained partially covered by chalky sediment. Therefore, a well-preserved sample, in which most microfossils were recognizable, and a poorly preserved sample, containing minor signs of lithification and generally poorly recognizable microfossils, were selected from the study material to assess the influence of initial preservation on the process. The well-preserved sample was split into 16 equal parts of ±40 mg. The poorly preserved sample, containing small lithified fragments, was split into 8 equal parts of ±40 mg. The sample splits were put in plastic cups filled with demineralized water and subjected to different durations of submersion in a Branson 2210 ultrasound bath at a frequency of 40 kHz. Afterwards the splits were dry-sieved, weighed and ultimately wet-sieved and weighed again.

The cleanness of the tests is assessed by observations on 200 randomly selected foraminifera. A specimen is considered clean if the test is free of sediment and sutures are clearly visible. Fragmentation is measured by the percentage of tests containing holes. To assess the taphonomic effect on different size fractions and for comparison with older literature, both the >63 µm and >125 µm fraction are evaluated. Removal of fine-grained material is calculated by the loss of mass >63 µm during each treatment step.

For both samples, the remaining mass decreases with increasing ultrasound duration (Fig. 1a). This is explained by removal of fine-grained material and by fragmentation of tests. Very little mass was lost after dry sieving and, although this increases with ultrasound duration, microfossils remained covered with sediment and cleanness improved only marginally. Subsequent wet sieving yielded significant cleanness improvements. After 18 min (1080 s) ultrasound exposure, the well-preserved sample had lost >20% of its mass. Fig. 1.

(a) Mass loss with increasing duration of ultrasonic treatment. (b) Percentage of fragmented planktic foraminifera. (c) Percentage of clean foraminifera. (d) SEM images of planktic foraminifera from the well-preserved sample for 0, 2, 8 and 18 min of ultrasound time. The close-ups show details of the surface texture of the planktic foraminifera at 0 and 18 min. (e) SEM images of planktic foraminifera from the poorly preserved sample for 2 and 8 min of ultrasound time. Scale bars in (d) and (e) delineate 100 and 10 µm for the close-ups.

Assessing the influence of ultrasound on test preservation has been attempted before with contradictory results, probably due to differences in starting material, methods and/or studied size fraction. Our preservation states after ultrasonic treatment and subsequent wet-sieving are controlled by a trade-off between cleanness and test damage. Longer ultrasound exposure yields generally cleaner foraminifera, which is intended, yet at the cost of a larger number of damaged foraminifera, which should be avoided for assemblage analyses. In our experiment, the optimal duration is 2–5 min, but in order to propose a standard procedure for ultrasound treatment, this should be further tested on a case-by-case basis.

In case ultrasound treatment is needed to improve the quality of a microfossil assemblage, determination of the required duration has to take into account the initial preservation state and the envisaged research goal. Fragile microfossils, such as most planktic foraminifera, require a shorter duration to avoid fragmentation. A study dealing mainly with benthic foraminifera may benefit from longer ultrasound treatment. When microfossils are fully covered with sediment, or when limited lithification is present, longer treatment times could also be advisable. A wide range of possible factors can diminish the initial preservation and, while some of these can be remediated by ultrasonic treatment, this is no catch-all method for easy improvements. A more in-depth study should try to assess more shell parameters on a generic or species level within different lithologies.