Direct assessment of stable isotope variability in planktonic foraminifera populations

Abstract

A large number of tests of three species of planktonic foraminifera, Globigerinoides sacculifer, Orbulina universa, and Pulleniatina obliquiloculata, were analyzed for stable oxygen and carbon isotopes in a western equatorial Pacific box core. Estimates of population variance were calculated using a statistical technique known as the jackknife. These estimates were used to generate sampling curves which provide the number of foraminiferal tests required in an isotopic analysis when precision and reliability criteria are specified.The jackknife results have important sampling implications: (1) a large number of foraminiferal tests are required to obtain statistically reliable isotope data. For example, given a machine precision (standard deviation) of 0.09in0, 417 tests (G. sacculifer, 355–420 μm) are required to obtain a δ18O precision of 0.1in0 with 90% reliability. (2) Variance is a function of test diameter, with greater variance being found in larger size fractions. (3) Variance is dependent on foraminiferal species. (4) Variance differences between glacial and interglacial samples are not significant. (5) In the size interval 600–900 μm, a systematic relationship is not seen between size and δ18O or δ13C. (6) Replicate analysis is a practical method of improving isotope data quality.Size- and species-related variance differences are probably related to a number of factors, including depth habitat, seasonal production of tests, species vertical migrations, and vital and/or symbiosis-caused effects. Knowledge of data variance permits error assessment in sediment “unmixing” analyses and optimization of spectral estimates.