Is the damage worth it? Testing handheld XRF as a non-destructive analytical tool for determining biogenic bone and tooth chemistry prior to destructive analyses

Isotopic analysis is destructive and requires that a specimen retains its original (biogenic) chemical composition. A specimen's relative abundance of calcium and phosphorous (Ca/P) or carbonate and phosphate (CO3/PO4) is often used to assess preservation. If a specimen's Ca/P or CO3/PO4 is similar to modern specimens, a specimen's isotopic composition may be biogenic. However, most methods for measuring these proxies are destructive. Moreover the relationships between Ca/P, CO3/PO4 and isotopic preservation are poorly established. In this study, we assessed the ability of handheld X-ray fluorescence (hXRF) to non-destructively evaluate a specimen's preservation by characterizing the calcium to phosphorous ratio (Ca/P). We first established that surface Ca/P (Ca/Psurface) for modern specimens was consistent with expectations for unaltered bone (1.3–2.3). Several specimens had slightly larger ratios, suggesting the currently accepted range may need to be expanded. Second, we tested the ability of Ca/Psurface to detect alteration using twenty Quaternary mammal teeth from Big Bone Lick, Kentucky. Ten specimens had Ca/Psurface between 1.3 and 2.3 and ten had larger ratios, suggesting alteration. Because most methods measure Ca/P in powder (Ca/Ppowder), we compared Ca/Psurface, Ca/Ppowder, and the enamel subsurface (Ca/Psubsurface). With two exceptions, Ca/Psubsurface and Ca/Ppowder were below 2.3, regardless of Ca/Psurface, suggesting that Ca/Ppowder and Ca/Psubsurface underestimate alteration. We next compared Ca/Psurface, CO3/PO4, and carbon (δ13C) and oxygen (δ18O) isotope values for the fossil teeth. Fourteen specimens were identified as altered or unaltered by both proxies, but six specimens only had one altered proxy. Specimens with both proxies altered had lower, less variable δ13C values than specimens with both proxies unaltered. Median δ18O values were similar between these groups. Individuals with altered Ca/Psurface but unaltered CO3/PO4 isotopically resembled specimens with both proxies altered. Conversely, specimens with unaltered Ca/Psurface and altered CO3/PO4 were similar to specimens with both proxies unaltered. Notably, all individuals with both proxies altered had relatively low δ13C values, including a horse and mammoth, which are normally considered grazers (and therefore should have higher δ13C values). These and other altered specimens may be isotopically compromised. Overall, our results suggest that Ca/Psurface is effective at detecting alteration non-destructively, quickly, and affordably, making it an attractive approach for analyzing unique specimens.