25Little is known of the earliest stages of soil ecosystem development on volcanic ash, and how 26 this process is affected by temperature. We studied the first three years of soil development in 27 a field-based mesocosm experiment, situated in different climates across Japan. Newly fallen, 28 sterilized volcanic ash from the Sakurajima volcano (Kyushu, Japan) was placed into pots 29 and positioned at six locations with mean annual temperatures ranging from -1.6 ℃ to 18.6 ℃.
30At 24 months into the experiment, C and N accumulation showed only a weak linear 31 correlation with temperature, but by 36 months there was a clear exponential relationship.
32This applied only to the top 2 cm of the developing soil, and was not apparent in the lower 33 part of the ash. We suggest that this acceleration in warmer climates relates to a positive 34 feedback involving bryophyte cover, which had become much denser by the third year in the 35 warmer sites. Surprisingly, the abundance of 16S rRNA gene copies of bacteria, fungi, 36 archaea -as well as ammonia oxidizersdid not increase from 12 months to 36 months, and 37 did not show any relationship to temperature, suggesting that input from plants is the major 38 factor in increasing C and N buildup in the soil. Overall it appears that temperature effects on 39 bryophyte cover buildup may be important in controlling the temperature relationship in soil 40 development on volcanic ash. 41 42 Weathering, Volcanic ash 44 Synnott, 1971). 70 It is also unclear what relationship, if any, the process of soil ecosystem development 71 shows to climate -especially temperature. Temperature is seen as a major controlling factor 72 in ecosystem establishment or recovery (Pastor and Post,1986; Vitousek and Farrington, 73 1997) as well as pedogenesis (Tsai et al., 2010), but it is unclear how early in the process it 74 becomes important. In this study we were interested in experimentally comparing the effect 75 of temperature on soil ecosystem development from volcanic ash, over the first three years of 76 primary succession. This study was a follow-on from Kerfahi et al. (2017), which discussed 77 the first two years of experimental soil ecosystem development. As we will report here, a 78 pattern was obtained when Year 3 was added to the dataset, with implications for 79 understanding the processes of ecosystem establishment on volcanic ash. 80 Furthermore, we report here on the inorganic nitrogen (N) dynamics at very early 81 stage of soil development. Nitrogen is not normally present in parent materials such as rock, 82 lava and volcanic ash in certain kinds of metasedimentary and metavolcanic rocks (Holloway 83 et al. , 1998). N deposition from the atmosphere is thought to be the primary source in the 84 earliest stages of microbial community development on volcanic ash, while microbial N 85 fixation eventually becomes more important.
86Our main working hypothesis here was that there would be a temperature effect on 87 the rate of soil ecosystem development (organic C, organic N, inorganic...