The development of a bioindicator system for soil acidity based on arthropod pH preferences.
A bioindicator system that supported the interpretation of acidity-related changes in soil chemistry was developed, using the tendencies of soil arthropods (springtails, mites, woodlice) to settle in a gradient of soil pH from 2 to 9. The system consisted of a circular walkway, the bottom of which was divided into 16 compartments filled with purified sand of varying pH. An extensive series of experiments were conducted to investigate the reproducibility of preference distributions observed for Orchesella cincta (Collembola) and Platynothrus peltifer (Cryptostigmata). There was a great variety of responses observed among 20 arthropod species tested. Based on the frequency distributions observed, each species was assigned a median preferred pH and an 'indicator value', to denote the specificity of the response. Many species had a broadly dispersed preference distribution, but the median preferred pH varied from 2.9 to 7.6. Three mites (Nothrus silvestris, Rhysotritia duplicata, and Odontocepheus elongatus) did not react to pH. The collembolan Folsomia candida showed a weak and variable response, but avoided pH 2. Three other Collembola (Orchesella flavescens, O. cincta, Lepidocyrtus cyaneus) and three isopods (Trichoniscus pusillus, Porcellio scaber, Philoscia muscorum) preferred the sub-neutral range. Species with a more or less pronounced preference for one or other end of the gradient were one collembolan (Tomocerus flavescens), two mites (Hypochthonius rufulus and Adoristes ovatus) and one isopod (Oniscus asellus), which were classified as acidophilous, and two Collembola (Isotoma notabilis, Entomobrya corticalis), two mites (Pelops occultus, Platynothrus peltifer) and one isopod (Armadillidium vulgare), which were classified as alkalophilous. The species Tomocerus minor showed a bimodal preference distribution. An 'arthropod acidity index' was proposed which allowed the median preferred pH of an arthropod community to be estimated from the indicator values, in conjunction with abundance scores in the field. Preliminary calculations showed that the index corresponded to differences between forests regarding soil pH, but the absolute pH indicated was one unit above litter pHs. When calibrated to field data, the proposed bioindicator system could be used in monitoring programmes for the analysis of ecological effects of long-term trends in soil acidity.