Rationale: Oxygen isotope ratios (δ 18 O values) of fish otoliths (ear bones) are valuable geochemical tracers of water conditions and thermal life history. Delta Smelt (Hypomesus transpacificus) are osmerid forage fish endemic to the San Francisco Estuary, California, USA, that are on the verge of extinction. These fish exhibit a complex life history that allows them to survive in a dynamic estuarine environment; however, a rapidly warming climate threatens this thermally sensitive species. Here we quantify the accuracy and precision of using δ 18 O values in otoliths to reconstruct the thermal life histories of Delta Smelt. Methods: Delta Smelt were reared for 360 days using three different water sources with different ambient δ 18 O water values (−8.75‰, −5.28‰, and −4.06‰) and different water temperatures (16.4°C, 16.7°C, 18.7°C, and 20.5°C). Samples were collected after 170 days (n = 28) and 360 days (n = 14) post-hatch. In situ δ 18 O values were measured from the core of the otolith to the dorsal edge using secondary ion mass spectrometry (SIMS) to reconstruct temporally resolved thermal life histories. Results: The δ 18 O otolith values for Delta Smelt varied as a linear inverse function of water temperature: 1000 ln α = 18.39 (±0.43, 1SE)(10 3 TK −1 ) − 34.56 (±1.49, 1SE) and δ 18 O otolith(VPDB) − δ 18 O water (VPDB) = 31.34(±0.09, 1SE) − 0.19(±0.01, 1SE) × T°C. When the ambient δ 18 O water value is known, this species-specific temperature-dependent oxygen isotope fractionation model facilitated the accurate (0.25°C) and precise (±0.37°C, 2σ) reconstruction of the water temperature experienced by the fish. In contrast, the use of existing general fractionation equations resulted in inaccurate temperature reconstructions. Conclusions: The species-specific δ 18 O otolith fractionation equation allowed for accurate and precise reconstructions of water temperatures experienced by Delta Smelt. Characterization of ambient δ 18 O water values remains a critical next step for reconstructing thermal life histories of wild Delta Smelt. This tool will provide new insights into habitat utilization, potential thermal refugia, and resilience to future warming for this critically endangered fish.

The acidic sulfate-rich waters of the Meridiani Planum region were potentially a habitable environment for iron-oxidizing bacteria on ancient Mars. If life existed in this ancient martian environment, jarosite minerals precipitating in these waters may record evidence of this biological activity. Since the Meridiani jarosite is thermodynamically stable at the martian surface, any biosignatures preserved in the jarosites may be readily available for analysis in the current surface sediments during the ongoing robotic exploration of Mars. However, thermal decomposition experiments indicate that organic compound detection of sediments containing jarosite may be challenging when using pyrolysis experiments; the instrument commonly used to assess organic matter in martian samples. So, here, we assess if the biogenicity of the Meridiani-type jarosites can be determined using complimentary spectroscopic techniques also utilized during the robotic exploration of Mars, including the upcoming ExoMars2020 rover mission. An abiotic jarosite, synthesized following established protocols, and a biological jarosite counterpart, derived from a microbial enrichment culture of Rio Tinto river sediments, were used to compare four spectroscopy techniques employed in the robotic exploration of Mars (Raman spectroscopy, mid-infrared (IR) spectroscopy, visible near-infrared reflectance (VNIR) spectroscopy and Mössbauer spectroscopy) to determine if the complimentary information obtained using these instruments can help elucidate the biological influence of Meridiani-type jarosites. Raman spectral differences might be due to the presence of unreacted reagents in the synthetic spectra and not biological contributions. Reflectance (IR/VNIR) spectra might exhibit minor organic absorption contributions, but are observed in both sample spectra, and do not represent a biosignature. Mössbauer spectra show minor differences in fit parameters that are related to crystal morphology and are unrelated to the biological (i.e., organic) component of the system. Results of this study suggest that the identification of biosignatures in Meridiani-type jarosites using the in situ robotic exploration on Mars may be possible but will be challenging. Our work provides additional insight into extraterrestrial biosignature detection and data interpretation for Mars exploration and indicates that sample return missions are likely required to unequivocally resolve the possible biogenicity of the Meridiani sediments or other jarosite-containing sediments.

The surname of the sixth author of Willmes et al 1 was incorrectly spelled in the original published article as "Christian Denny". The correct name is "Christian Denney".

Analytical protocols for SHRIMP‐SI oxygen isotope analysis (δ18O) of a suite of zircon reference materials (RMs) are presented. Data reduction involved a robust estimate of uncertainties associated with the individual spot as well as for groups where the spot data are combined. The repeatability of δ18O measurements is dependent on both the analytical conditions and the choice of the primary reference material. Under optimised conditions, repeatability was often better than 0.4‰ (2s) allowing sample uncertainties to be obtained to better than 0.2‰ (at 95% confidence limit). Single spot uncertainty combined the within‐spot precision with the scatter associated with repeated measurements of the primary zircon reference material during a measurement session. The uncertainty for individual spots measured under optimised conditions was between 0.3 and 0.4‰ (at 95% confidence). The analytical protocols described were used to assess a variety of zircon RMs that have been used for geochronology and for which laser fluorination oxygen isotope data are available (Temora 2, FC1, R33, QGNG and Plešovice), as well as zircons that have been used as RMs for trace element or other types of determination (Mud Tank, Monastery, 91500, AS57, AS3, KIM‐5, OG1, SL13, CZ3 and several other Sri Lankan zircons). Repeated analyses over nine sessions and seven different mounts show agreement within analytical uncertainty for Temora 2, FC1, R33, QGNG, Plešovice and 91500, when normalised to Mud Tank. For existing ion microprobe mounts with these materials, an appropriate δ18O can be determined. However, care should be taken when using zircons from the Duluth Complex (i.e., FC1, AS57 and AS3) as reference materials as our data indicated an excess scatter on δ18O values associated with low‐U zircon grains.