Can wet summers trigger permafrost collapse at a Siberian lowland tundra site?

In recent decades, tundra shrub expansion has been observed in many places in response to Arctic climate warming. We question whether this shrub expansion is the only direction in which tundra landscapes will change. At our Siberian lowland tundra research site we observe both local shrub dominance and shrub drowning when conditions become too wet due to local permafrost collapse. In lowland tundra, which is poorly drained and underlain by ice-rich permafrost, shrub decline due to permafrost degradation could be a likely alternative for the widely assumed shrub expansion.
In 2007 we started a Betula nana shrub removal experiment at a Northeast-Siberian tundra site to investigate how shrub expansion would influence the summer thawing of permafrost. The removal of the shrub part of the vegetation initiated thawing of per­mafrost , resulting in collapse of the originally elevated shrub patches into waterlogged depression within five years (Nauta et al. 2015). These results demonstrate the extreme sensitivity of these lowland tundra eco- systems to perturbations. It was not only within our experiment plots that the permafrost collapsed. We observed many 'natural' thaw ponds in the study area for which it often is unclear what caused the abrupt thawing.
We think that wet summers, such as in 2011, could have been a trigger for the local permafrost collapse. Niue years of measurements in the unchanged control plots of the removal experiment showed that the thawing depth was not largest in the warmest summer, but in the wettest summer (2011), implying an important role for precipitation in this ecosystem. Using a navel application of dendrochronological methods to drowned shrubs, we assessed whether thaw pond development and associated shrub drowning can be traced back to recent climatic changes. Preliminary results indicate that most of the studied thaw ponds developed recently.
If a future warmer and wetter climate can more frequently trigger such local permafrost collapse, methane-emitting wetlands and thermokarst ponds would expand in the Siberian lowland tundra landscape, which contrasts with the widely assumed shrub expansion.