Inter- and intra-observer variation in phytolith morphometry.

Area of Science:

• Archaeobotany
• Palaeoecology
• Geometric Morphometrics

Background:

• Geometric morphometrics is crucial for identifying plant opal phytoliths in archaeobotany and palaeoecology.
• Morphometric data from archaeological phytoliths aids taxonomic attribution when compared to reference material.
• Observer variation in phytolith morphometrics is an understudied source of error and potential misidentification.

Purpose of the Study:

• To investigate inter- and intra-observer variation in phytolith morphometrics.
• To identify factors contributing to variation in phytolith measurements.
• To provide recommendations for minimizing observer error in phytolith analysis.

Main Methods:

• Eight researchers measured three phytolith morphotypes (Bilobate, Bulliform flabellate, Elongate dendritic) from 50 samples each, three times.
• Measurements involved digitizing phytolith outlines from photographs and analyzing them with open-source morphometric software.
• Data collected included 17 size and shape variables for each phytolith.

Main Results:

• Inter-observer variation ranged from 0% to 23% difference from the mean.
• Intra-observer variation ranged from 0% to 9% difference from the mean.
• Variation was generally higher among inexperienced researchers, with scaling errors being a major cause.

Conclusions:

• Inter- and intra-observer variation can be significantly reduced through standardized instructions and training.
• The International Committee for Phytolith Morphometrics (ICPM) provides recommendations to minimize variation.
• Standardized guidelines are essential for accurate, consistent, and comparable phytolith morphometric analyses until automatic data collection is widespread.