Assessment of bias in carbon isotope composition of organic leaf matter due to pre‐analysis milling methods

Abstract

Rationale: Stable isotope analysis of leaf material has many applications including assessment of plant water-use efficiency and paleoclimatology. To facilitate interpretations of small shifts in the carbon isotope composition (δ 13 C) of leaves, accurate and repeatable results are required. Pre-sample homogenisation is essential to ensure a representative sample is analysed, but can also introduce error. Methods: We investigate how different grinding methods (freezer-milling and ball-milling) affect the carbon content and δ 13 C of tree leaves from a wetland in Queensland, Australia, commenting on how increased temperature, sample contamination, sample loss, or poor homogenisation may impact results. Results: No alteration of leaf δ 13 C is observed due to different milling methods, although there may be a significant increase in %C of samples processed using ball-milling. Conclusions: We suggest %C variability is possibly due to contamination from abraded plastic vials or insufficient homogenisation during ball-milling, with no significant impact on δ 13 C. Overall, we suggest that intermittent ball-milling may be the best solution to reduce costs, preparation time and use of liquid nitrogen, aiming to achieve complete homogenisation using the shortest possible duration of milling.

No evidence for effects of mill‐grinding on δ13C, δ15N and δ34S values in different marine taxa

“…However, these values do not overlap with some ecologically important primary producers, e.g., C 4 plants or seagrass 24 . Therefore, any contamination of samples by plastic could potentially result in isotopic noise, leading to misinterpretation of results and any subsequent conclusions 9,17,22 …”

“…A range of other homogenization techniques do exist, including freezer milling. However, access to such specialist equipment 17 is less common than for the pestle and mortars used in hand grinding or the bead‐beaters which are commonly available in less specialized (i.e., not stable isotope) biology laboratories.…”

“…A small number of existing studies have compared the use of different homogenization treatments. However, these have largely focused on potential biases of δ 13 C values in a limited set of sample types (mostly terrestrial) 17,22 . Here we examine potential preparation effects in a diverse range of taxa that make major contributions to coastal marine food webs, and that are commonly sampled by stable isotope ecologists working in these habitats.…”

“…15,16 For instance, elemental and isotopic compositions (e.g., C, N and S) can vary considerably within a given sample. [17][18][19] Typically, solid samples for stable isotope analysis are homogenized by grinding them to fine powders. However, if grinding results in insufficient homogenization of particularly intractable matrices (e.g., macroalgae which can be resistant to homogenization), this could potentially effect stable isotope values considering the variability in elemental or isotopic composition reported within individuals, as seen in δ 34 S values in different parts of Posidonia oceanica seagrass, which reflect exposure to different sulfide concentrations in leaves, rhizomes and roots.…”

“…24 Therefore, any contamination of samples by plastic could potentially result in isotopic noise, leading to misinterpretation of results and any subsequent conclusions. 9,17,22 A small number of existing studies have compared the use of different homogenization treatments. However, these have largely focused on potential biases of δ 13 C values in a limited set of sample types (mostly terrestrial).…”