This paper examines variation of phytoliths in leaves of corn (Zea maysL.) as part of an ongoing project to establish accurate, reliable, phytolith "signatures." Factors investigated include leaf part, individual plant, leaf node, geographic location, and corn variety. Counts of 100 phytoliths were made on 24 separate samples, classifying shapes into dumbbell, cross, and irregular categories. Resulting assemblages were compared using chi-square andNANOVAtest statistics. Results indicate that different leaf parts and corn varieties significantly affect phytolith assemblages. Base, midsection, and tip of leaf blades consistently produce different assemblages; two North Dakota native corn varieties also produce different assemblages (when part variation is eliminated). Variation due to leaf node and individual plant is not consistent; variation due to geographic location is generally not significant. It is suggested that researchers use entire leaf blade sections as samples in phytolith analysis. This sampling design will incorporate variation exhibited within the leaf blade and avoid characterization of plants by inaccurate phytolith assemblages.
展开▼
