Vibrational signals as indicators of soil fauna health? : A novel approach to environmental monitoring of ants

Sammanfattning: Soil fauna provides processes of crucial importance for ecosystem functions, but our ability to observe their actions often depend on destructive methods where the integrity of the studied environment (the soil) is compromised. In this study, I develop and test a new generation of environmental monitoring tools that utilize vibrations made by soil macrofauna to inform about their performance and health. Three hypotheses were tested on forest ants (Formica sp.): (i) vibrations on a naturally occurring substrate can be used to measure the activity of ants, (ii) the vibrational signature of ants can inform us about the ants’ health, and (iii) behavior (locomotory activity, foraging etc.) of ants is correlated to specific vibrational signal characteristics. Vibrational signals from ant sub-populations (5 individuals/sample) were recorded on natural substrates (leaves) before and after exposure to sub-lethal levels of a contaminant with known negative effects on ants (imidacloprid). Activity was successfully detected from the inferred vibration oscillograms. However, neither the number of vibrational signals captured, or the signal characteristics changed after imidacloprid exposure, indicating that either the exposure was too low to generate a change in behavior or that the technique was unable to detect subtle changes in behavior. Signals of short duration and amplitude were generated by locomotory activity, but their intensity was dependent on the distance to the vibrometers laser focus point. Longer signal segments were either continuous, random, or rhythmic in their distribution. I suggest that these signals are a result of methodological artifacts, movement of multiple ants, and stridulation/drumming on the substrate, respectively. My findings suggest that this advanced monitoring tool is sensitive enough to capture activity from only a few ants, but the characteristics and number of captured signals are strongly affected by the measured substrate