Historical variation in the normalized difference vegetation index compared with soil moisture in a taiga forest ecosystem in northeastern Siberia

Abstract. The taiga ecosystem in northeastern Siberia, a nitrogen-limited ecosystem on permafrost with a dry climate, changed during the extreme wet event in 2007. We investigated the normalized difference vegetation index (NDVI) as a satellite-derived proxy for needle production and compared it with ecosystem parameters such as soil moisture water equivalent (SWE), larch foliar C/N ratio, δ13C and δ15N, and ring width index (RWI) at the Spasskaya Pad Experimental Forest Station in Russia for the period from 1999 to 2019. Historical variations in NDVI showed a large difference between typical larch forest (unaffected) and the sites affected by the extreme wet event in 2007 because of high tree mortality at affected sites under extremely high SWE and waterlogging, resulting in a decrease in NDVI, although there was no difference in the NDVI between typical larch forest and affected sites before the wet event. Before 2007, the NDVI in a typical larch forest showed a positive correlation with SWE and a negative correlation with foliar C/N. These results indicate that not only the water availability (high SWE) in the previous summer and current June but also the soil N availability likely increased needle production. NDVI was also positively correlated with RWI, resulting from similar factors controlling them. However, after the wet event, NDVI was negatively correlated with SWE, while NDVI showed a negative correlation with foliar C/N. These results indicate that after the wet event, high soil moisture availability decreased needle production, which may have resulted from lower N availability. Foliar δ15N was positively correlated with NDVI before 2007, but foliar δ15N decreased after the wet event. This result suggests damage to roots and/or changes in soil N dynamics due to extremely high soil moisture. As a dry forest ecosystem, taiga in northeastern Siberia is affected not only by temperature-induced drought but also by high soil moisture (led by extreme wet events) and nitrogen dynamics.