Spatiotemporal variability in precipitation-growth association of Betula nana in the Siberian lowland tundra

Shrubs are expanding across a warming Arctic, evident from range expansion and increases in biomass, stature and cover. This influences numerous aspects of Arctic ecosystems. While shrub growth is generally positively associated with summer temperature, tundra ecosystems are characterised by abiotic gradients on small spatial scales (metres), and the Arctic climate and its year-to-year variability are changing rapidly. Hence, it is often unclear to what extent climate-growth associations are scalable to future climate scenarios and across environmental gradients within ecosystems. Here, we investigate the stability of climate–growth associations of Arctic dwarf shrubs across small-scale (metre to kilometre) topographic gradients and decadal timescales. We constructed ring width series (1974–2018) for a common Arctic dwarf shrub (Betula nana) for three representative types of subsites in the Siberian lowland tundra: higher elevation, lower elevation and thermokarst-affected (thaw ponds) terrain. We quantified decadal variability in climate–growth associations across subsites using partial least squares regression and a moving window approach. We found consistently positive association of shrub radial growth with summer temperature, but substantial spatial and temporal variability in precipitation response. Association of shrub growth with summer rainfall increased in recent decades. Shrubs on elevated sites showed particularly strong response to rainfall following drier periods, and a negative association with recent snowfall extremes. Shrubs sampled from thaw ponds showed strong positive association with rainfall, followed by high shrub mortality after an extremely wet summer. This likely resulted from waterlogging due to thermokarst. Synthesis. Our findings imply that the response of shrub growth to changes in Arctic precipitation regimes is regulated by (i) macro- (kilometre-scale) and micro-topographical (metre-scale) gradients, (ii) colimitation between temperature and moisture and (iii) potentially nonlinear responses to precipitation extremes. This suggests that the scalability of precipitation-growth relationships for Arctic shrubs across dynamic tundra landscapes and future climate scenarios is limited. We recommend that future climate–growth studies on Arctic tundra shrubs simulate future precipitation changes across spatial gradients and include detailed microsite and shrub physiological monitoring.