An integrated leaf trait analysis of two Paleogene leaf floras

OBJECTIVES: This study presents the Integrated Leaf Trait Analysis (ILTA), a workflow for the combined application of methodologies in leaf trait and insect herbivory analyses on fossil dicot leaf assemblages. The objectives were (1) to record the leaf morphological variability, (2) to describe the herbivory pattern on fossil leaves, (3) to explore relations between leaf morphological trait combination types (TCTs), quantitative leaf traits, and other plant characteristics (e.g., phenology), and (4) to explore relations of leaf traits and insect herbivory. MATERIAL AND METHODS: The leaves of the early Oligocene floras Seifhennersdorf (Saxony, Germany) and Suletice-Berand (Ústí nad Labem Region, Czech Republic) were analyzed. The TCT approach was used to record the leaf morphological patterns. Metrics based on damage types on leaves were used to describe the kind and extent of insect herbivory. The leaf assemblages were characterized quantitatively (e.g., leaf area and leaf mass per area (LM(A))) based on subsamples of 400 leaves per site. Multivariate analyses were performed to explore trait variations. RESULTS: In Seifhennersdorf, toothed leaves of TCT F from deciduous fossil-species are most frequent. The flora of Suletice-Berand is dominated by evergreen fossil-species, which is reflected by the occurrence of toothed and untoothed leaves with closed secondary venation types (TCTs A or E). Significant differences are observed for mean leaf area and LM(A), with larger leaves tending to lower LM(A) in Seifhennersdorf and smaller leaves tending to higher LM(A) in Suletice-Berand. The frequency and richness of damage types are significantly higher in Suletice-Berand than in Seifhennersdorf. In Seifhennersdorf, the evidence of damage types is highest on deciduous fossil-species, whereas it is highest on evergreen fossil-species in Suletice-Berand. Overall, insect herbivory tends to be more frequently to occur on toothed leaves (TCTs E, F, and P) that are of low LM(A). The frequency, richness, and occurrence of damage types vary among fossil-species with similar phenology and TCT. In general, they are highest on leaves of abundant fossil-species. DISCUSSION: TCTs reflect the diversity and abundance of leaf architectural types of fossil floras. Differences in TCT proportions and quantitative leaf traits may be consistent with local variations in the proportion of broad-leaved deciduous and evergreen elements in the ecotonal vegetation of the early Oligocene. A correlation between leaf size, LM(A,) and fossil-species indicates that trait variations are partly dependent on the taxonomic composition. Leaf morphology or TCTs itself cannot explain the difference in insect herbivory on leaves. It is a more complex relationship where leaf morphology, LM(A), phenology, and taxonomic affiliation are crucial.