Differences in perception and judgment and their influences on knowledge are being evaluated in many domains of natural sciences and beyond. Among them, taxonomic uncertainty creates a known, but mostly ignored, identification bias in ecological studies. This bias undermines reproducibility, which is especially critical in comparisons among studies and meta-analyses.
Testate amoebae (TA) are commonly used as proxies for palaeoenvironmental reconstructions and as bioindicators in peatlands and lakes. Such work is always based on morphological identification, which for many taxa is challenging due to the lack of clear morphological criteria and identification keys.
We conducted a survey within the TA research community, using microscopy images of TA observed in surface Sphagnum and peat samples. These images were shared with all research groups working with TA. Observers with varying degrees of expertise participated, and broad descriptive data about each participant was documented. Additionally, we recorded self-evaluated level of confidence for each identification. We hypothesized that TA taxonomic identification bias would affect estimates of diversity and ecological interpretation including applications such as bioindication.
This is the first attempt to simultaneously include all the abovementioned metrics in one analysis. Notably, no expert judgement from our side or from any previously made research was used. Our results create a fundament for work on methodological guidelines and exploration of approaches allowing us to overcome identification bias. This will help make TA-based ecological studies more robust.
