First absolute seasonal temperature estimates for greenhouse climate from clumped isotopes in bivalve shells

Abstract

Seasonal variability in sea surface temperatures plays a fundamental role in climate dynamics and species distribution. As such, it is essential to better understand seasonal variability in climates of the past. Previous reconstructions of seasonality in deep time are poorly constrained, relying on controversial assumptions such as estimates of seawater composition and neglect seasonal bias. This work presents the first absolute seasonal temperature reconstructions based on clumped isotope measurements in bivalve shells which, critically, do not rely on these assumptions. Our new approach reconstructs highly precise higher mid-latitude (~50°N) monthly temperatures from individual oyster and rudist shells of the Campanian (78 million years ago) greenhouse period (15—27 °C seasonal range). Our analysis demonstrates that seasonal bias and previous assumptions about sea water oxygen isotope composition can lead to highly inaccurate temperature reconstructions, distorting our understanding of the behavior of greenhouse climates and our ability to model them. Our results agree with fully coupled climate model simulations showing greenhouse climates outside the tropics were warmer and more seasonal than previously thought.

The potential of marine bivalve Spisula sachalinensis as a marine temperature record

“…Clumped isotope measurements as a palaeoclimatic proxy within Spisula sachalinensis may be a means to circumvent some of the issues identified here. This approach has been successfully used to provide seasonal climate reconstructions from Cretaceous bivalves (de Winter et al, 2020), and has the benefit of being independent from estimations of past δ 18 O seawater (Eiler, 2007;Henkes et al, 2013), which we have already noted among the confounding factors of temperature interpretation. This will be especially relevant for prehistoric and deep-time bivalves where historical records of seawater composition can't be relied upon, but could also be useful in cases of extreme seasonal variation in δ 18 O seawater which can be hard to incorporate into palaeoclimate models.…”