Mg/Ca ratios in Californian mussels

After meeting Christopher Jazwa (Assistant Professor, University of Nevada) at the SAAs earlier this year, he sent me some of his Mytilus californianus shells to play with and test our LIBS system on.

The shells are from Santa Rosa Island, which is known for its abundance of shell middens (albeit less entertaining than the middens on San Miguel Island). Santa Rosa middens and their shells are a massive source for climate data ranging back into the Mid-Holocene and provide first hand insights into the human exploitation of marine shells.

Christopher analyses the shell's geochemical properties, because these can change as a result of environmental change. The shells that Chris sent over were modern shells that he previously analysed using stable isotope measurements (δ18O) to see what environmental change generally looks like within the records of Californian musses from Santa Rosa.

Now we wanted to see, what the ratio of Magnesium to Calcium in the shell can tell us.

Mg/Ca ratios are regularly tested to see whether they can be used for reconstructing water temperatures, but are also often shown to be not very reliable. So mapping the Mg/Ca ratios using LIBS would give us pretty a good idea of what's going on in this case. Our new setup uses a combination of cleaning shots and measurement shots. Cleaning shots that take off ~0.1µg from the surface and expose fresh carbonate, while measurement shots actually measure Magnesium and Calcium concentrations, using our trusted peaks of Mg and calcium ions with their respective wavelengths of 279 nm and Ca ions at 315 nm. The intensity of both peaks is based on the intensity of emitted photons with the above wavelengths. With each shot we create a plasma that emits photons based on the plasma's elemental composition. That way we can get a linear correlation between the peaks' intensities and the actual Mg/Ca ratio within the material.

---

​Looking at the shells in general, there are two layers of the shell that grow pretty much in sync. One is made of calcite (coloured) and one is made of aragonite (black). The reason why the aragonitic (or inner) layer is black, is because it contains a much lower ratio of Mg to Ca and can pretty easily be identified this way.

The calcitic (or outer) layer of the shell has Mg/Ca ratios that range between 0.05 and 0.25.

Another thing that became clear is that all shells are different. These differences range from trivial things, like the holes from Christopher's earlier drilling for carbonate samples in January's shell #1, to less trivial things like growth interruptions and mixing of calcite and aragonite layers in May's shells #2 and #3.

The inconsistent way that these layers are mixed, makes them very difficult to analyse and to track changes in element concentration through time (so I didn't ).

The Mg/Ca ratio is mostly consistent within growth increments, which suggests that whatever controls the ratio is doing this at the point in time that the growth specific increments represent. This is different from the high and low ratios we find in the calcite and aragonite parts which grew at the same time, because their respective mineralogies are already biased towards one way or the other.

While all shells show some kind of banding patterns with high and low Mg/Ca ratios, these bands are not consistently distinct in all specimens. For comparison, there is map of a shell further below, with very distinct patterns from Christopher's March collection.

How come this distinctiveness is not always the case? - I don't know.

Could this be age related, since bands are getting gradually more distinct in this shell? - We'll need more older shells to determine this.

What could the Mg/Ca increases actually mean? - Tentatively, it could be argued that they are seasonal patterns that happen every year and high Mg/Ca ratios in one shell relate to similar times in another shell (stroked lines illustrate this). Sadly the inconsistency of the patterns, especially in shells from May, makes this a somewhat useless bit of information, which can't be relied on.

More importantly, the concept that Mg/Ca is mostly temperature related is seemingly not the case in our samples, because a) the factors behind the distortion of patterns are clearly in control of Mg/Ca ratios and dominate the record and b) the Mg/Ca ratios measured at the shell edges of January shells have a higher Mg/Ca ratio than the May shells, where water temperature (and thus Mg/Ca ratios) should be warmer.

This preliminary data leaves us with more questions to solve next:

• Are the distinctly high-ratio bands related to seasonal change?

• What causes the higher amounts of Mg/Ca in those bands (seasonal or structural factors)?

• Can the banding patterns be used to guide sampling for stable isotope analysis?

• What causes irregularities or the lack of them?

• Can these causes be anticipated by distinct exterior features of the shell?

• Are irregularities more prominent in younger shells/younger parts of old shells?

To answer this, we'll need many more shells to find common features and additional contextual information to characterise them.

Until soon,

Niklas