Evaluation of measurement error in rotational mounting of larval Rhipicephalus (Acari: Ixodida: Ixodidae) species in geometric morphometrics.

Geometric morphometric analysis, in combination with modern computational power, allows for precise measurements of morphological shape to study even minimal variation. One key challenge, however, is 2D photographs which serve as proxies for 3D structures that introduce measurement error into a dataset by omitting variability in the Z dimension. The ease of application of 2D proxies means we should optimize them, rather than limit studies by enforcing expensive or complex protocols. This may be done by explicitly measuring and reporting error. A rotational mounting surface should be capable of reducing measurement error associated with positioning (rotation). This study empirically quantified positioning error in basis capituli of three closely related Rhipicephalus species larvae. It investigates symmetrical and asymmetrical components of variation, in addition to the number of replicates required to obtain confident results. Asymmetry showed rotational error along the roll axis (side to side) which had a large effect that confounded species variation. In contrast, symmetrical variation showed much less rotational error, and only along a pitch axis (forwards and backwards) that was much smaller than species variation. Some systematic error was noted in roll axis variation that caused the left sides of specimens to be slightly raised. However, this error was small, and was avoidable when considering symmetrical variation. Confidence of differences in multivariate means between groups increased drastically with three replicates (15%), and slightly more with five (5%). This study demonstrates that variation at and below the individual level is out of scope for rotational mounting, and will provide unreliable results. However, this mounting method is validated for questions pertaining to species variation and taxonomy. [ABSTRACT FROM AUTHOR]