Soil Ecology : Long-term development of Collembola communities on a former rubble deposit

Eighteen sampling campaigns were evaluated for investigating the long-term secondary succession of Collembola communities in relation to climate change and vegetation and in comparison with Gamasina communities at two different managed experimental plots. One plot (SUC) was left undisturbed for natural secondary succession, the other plot (REC) was recultivated with rotary-tilling and sowing of grass. This experiment was done on a former rubble deposit over 20 years and samplings took place from 1980 to 2000. Selected sampling campaigns, the first three years (1980, 1981, 1982), the last three years (1998, 1999, 2000) and in -between with an interval of four years (1986, 1990, 1994), were evaluated and studied between 2014 and 2017. In total, 14,742 Collembola individuals were counted and identified by the author. Those consisted of 33 species of which 25 species were identified by the author herself. The most dominant species were: Mesaphorura krausbaueri, Parisotoma notabilis and Isotomodes productus. The analysis of Collembola from this collection posed great challenges in clearing the specimens stored in 70% ethanol for 15-32 years for microscopical inspection. None of the five media with different clearing properties tested for permanent slide mounts cleared the old specimens sufficiently, therefore, pretreatment procedures had to be applied. As such, to achieve good clearing of old ethanol collection material, it is recommended to incubate it in DNA extraction buffer solution such as SNET plus proteinase K and to mount it in Marc Andre 2 for best optical results. Long-term successional development of Collembola was investigated at the species level with calculation of abundance (ind/m2), dominance, diversity indices, presence-absence and dominance similarity and species turnover rate. Successional stages of Collembola species in both experimental plots were identified. Ecological groups were also distinguished for Collembola species. Long-term succession of Collembola was related to the changes of climatic variables and vegetation in the studied area over 20 years. This relationship was analysed and modeled applying canonical correspondence analysis, linear regression analysis and average linkage clustering for the Collembola abundance data. The model reveals four distinct climatic groups of evaluated data and shows a significant influence of soil temperature on the long-term successional development of Collembola species. No evidence was found for the correlation of Collembola species abundance with vegetation covers. Thus, they are both either independent from each other or at least not related directly. A comparative study was performed to study the synchronicity of the successional development of Collembola, Gamasina and plant species. This was also compared between two experimental plots. To compare the abundance data from Collembola with Gamasina, the same calculations were performed for successional development of Gamasina species, and the model of successional stages and the structure of ecological groups were also provided for Gamasina species. The successional synchrony among the three organism Groups was evaluated and modeled with detrended correspondence analysis. Redundancy analysis and linear regression revealed models for the evaluation of the influence of environmental variables on the successional development of investigated microarthropod groups. Again, soil temperature was proved as a significant factor for both groups, and pH was significantly influential in the longterm development of Gamasina species. Different management in SUC and REC played an important role in successional changes of the three organism groups; the difference between two plots was observable especially in Gamasina. Overall, Collembola community was more similar between both plots and Gamasina community was generally more diverse. Replacement of species during a successional study is mainly related to time i.e. their temporal shifts. Moreover, some environmental variables may influence this process effectively and significantly; in particular soil temperature and pH are determined as such important factors in this study