Contact solid-phase microextraction with uncoated glass and polydimethylsiloxane-coated fibers versus solvent sampling for the determination of hydrocarbons in adhesion secretions of Madagascar hissing cockroaches Gromphadorrhina portentosa (Blattodea) by

• Insect hydrocarbons of tarsal adhesion secretions of cockroaches analyzed by GC-MS.
• Contact SPME with uncoated glass fiber compared to polydimethylsiloxane PDMS fiber.
• Molecular hydrocarbon patterns sampled with glass and PDMS fiber are similar.
• Greatly simplified in vivo sampling procedure with glass fiber shows better precision.
• Hydrocarbons on tarsi less abundant but qualitatively similar to profiles on tibiae.

Molecular profiles of adhesion secretions of Gromphadorrhina portentosa (Madagascar hissing cockroach, Blattodea) were investigated by gas chromatography mass spectrometry with particular focus on a comprehensive analysis of linear and branched hydrocarbons. For this purpose, secretions from the tarsi (feet), possibly contributing to adhesion on smooth surfaces, and control samples taken from the tibiae (lower legs), which contain general cuticular hydrocarbons that are supposed to be not involved in the biological adhesion function, were analyzed and their molecular fingerprints compared. A major analytical difficulty in such a study constitutes the representative, spatially controlled, precise and reproducible sampling from a living insect as well as the minute quantities of insect secretions on both tarsi and tibiae. Thus, three different in vivo sampling methods were compared in terms of sampling reproducibility and extraction efficiency by replicate measurement of samples from tarsi and tibiae. While contact solid-phase microextraction (SPME) with a polydimethylsiloxane (PDMS) fiber showed higher peak intensities, a self-made uncoated glass fiber had the best repeatability in contact-SPME sampling. Chromatographic profiles of these two contact-SPME sampling methods were statistically not significantly different. Inter-individual variances were larger than potentially existing minor differences in molecular patterns of distinct sampling methods. Sampling by solvent extraction was time consuming, showed lower sensitivities and was less reproducible. In general, sampling by contact-SPME with a cheap glass fiber turned out to be a viable alternative to PDMS-SPME sampling. Hydrocarbon patterns of the tarsal adhesion secretions were qualitatively similar to those of epicuticular hydrocarbon profiles of the tibiae. However, hydrocarbons were in general less abundant in tarsal secretions than secretions from tibiae.