INSTITUTO ARGENTINO DE NIVOLOGIA, GLACIOLOGIA Y CIENCIAS AMBIENTALES
Unidad Ejecutora - UE
congresos y reuniones científicas
Chemometrics in Palaeobotany and its chemotaxonomic implications: a case study of neuropterid foliage (Pennsylvanian, Sydney Coalfield), Canada.
D`ANGELO, J A; ZODROW, E L; CAMÍ, G; RAMIREZ, D; MAMMANA, S
Simposio; XV Simposio Argentino de Paleobotánica y Palinología; 2012
Chemometrics combines chemical measurements (e.g., spectroscopic data) and mathematical or statistical methods (e.g., multivariate analysis). Spectroscopic data are unique because they are both multivariate in nature (several hundred variables) and highly correlated. Multivariate techniques (e.g., principal component analysis -PCA) in combination with spectroscopic data (e.g., Fourier Transform Infrared-FTIR) are powerful methods for identifying dominant data patterns (e.g., groups or trends). This case study illustrates the capabilities of chemometrics applied to FTIR data obtained for coalified compressions of several neuropterid taxa (Pennsylvanian, Sydney Coalfield, Canada). Neuropterids represent one of the most common and widespread types of foliage in Carboniferous time. However, they correspond to a grouping of many genera belonging to various families. Those genera are distinguished based on character sets including frond architecture, pinnule and cuticular morphologies, which are not equally known for each group. This makes it impossible to list the Carboniferous taxa unequivocally belonging to the neuropterids. Using semi-quantitative FTIR and PCA data evaluations, 98 coalified specimens of 9 neuropterid taxa were investigated to assess whether their chemical compositions can be used for chemotaxonomic purposes. PCA results successfully identified groups related to neuropterid taxa traditionally defined by morphology e.g., Macroneuropteris scheuchzeri (Hoffmann) Cleal et al. 1990 and Linopteris obliqua (Bunbury 1847) emend. Zodrow et al. 2007. Furthermore, some non-neuropterid specimens, i.e., Alethopteris pseudograndinioides Zodrow et Cleal 1998 could clearly be differentiated from the neuropterids. These findings underpin the potential for chemometrics (i.e., FTIR followed by PCA) to characterize neuropterid taxa and to differentiate between neuropterids and non-neuropterids.