The serial test dissection and sieve fraction methods for determining the pattern of size-related change in oxygen and carbon isotopic ratios are compared using four Late Cretaceous planktic foraminifer species (Racemiguembelina fructicosa, Planoglobulina acervulinoides, Planoglobulina multicamerata, and Pseudoguembelina palpebra) from a subtropical site in the North Atlantic (DSDP Hole 390A). Despite the extra labor required, we identify several clear advantages of the dissection method, including: (1) it provides a means of obtaining size-dependent changes in isotopic signatures that are unequivocally ontogenetic, whereas isotopic variation observed from sieve-separated size fractions could be ontogenetic or ecotypic; (2) the taxonomic identity of smaller sized specimens using the dissection method is unequivocal, whereas species identification is increasingly ambiguous in smaller size fractions using the sieve method; (3) it reveals a greater total range and a greater complexity in the pattern of ontogenetic change in stable isotopic values, whereas the sieve method averages the isotopic signal across the entire ontogenetic range preserved within the whole tests that are used. Our results from serial dissections demonstrate that among the species analyzed, R. fructicosa and P. acervulinoides yield relatively negative adult delta18O values, a large size-related change in delta13C values (1.32 and 2.05‰, respectively), and virtually no correlation between size-related delta13C and delta18O values. On this basis we suggest that these were photosymbiotic species that inhabited relatively shallow surface waters. Evidence for photosymbiosis is not as compelling for P. palpebra, as this species yields a 1.06‰ shift in delta13C and relatively negative delta18O values in adult chambers, but much stronger correlation between size-related delta13C and delta18O values (r square = 0.4) than in R. fructicosa and P. acervulinoides. Planoglobulina multicamerata yields the most positive adult delta18O values of the species studied, a strong covariance between size-related delta13C and delta18O values (r square = 0.77), and a 0.97‰ shift in d13C composition during ontogeny. We conclude that this species lacked photosymbionts and migrated to a deeper surface water paleohabitat as it increased in size. Single specimen analyses of tightly constrained size fractions reveal a high degree of intraspecific variation. Delta13C and delta18O values vary by up to 0.70 and 0.28‰ in R. fructicosa, 1.41 and 0.80‰ in P. acervulinoides, 0.66 and 0.82‰ in P. palpebra, and 0.18 and 0.33‰ in P. multicamerata, respectively. Such a range of isotopic variation has been observed in modern day planktic foraminifer assemblages, and likely results from growth of individuals during different phases of the seasonal cycle and/or the kinetic effect of intraspecific variation in shell calcification rates. As suggested by other investigators, large sample sizes should be analyzed to provide the most reliable correlation of stable isotopic stratigraphic records.
