Plant cover changes drive plant and soil carbon pool responses in High Arctic dry heath exposed to decades of experimentally increased summer rain and nutrient addition

With accelerating climate change, higher summer rainfall and warmer soils are expected for high Arctic ecosystems. Yet, how increased inputs of moisture and nutrients to soils will affect plant composition and ecosystem C storage in these arid, low-productivity ecosystems remains unclear. We utilised a long-term experiment in a dry shrub heath tundra in Zackenberg, NE Greenland, in which N and P availability was increased and precipitation doubled experimentally every summer for 25 years. We determined soil and vegetation C pools, plant cover and leaf chemistry, and ecosystem CO2 fluxes were measured over three weeks in peak growing season. Watering increased the cover of graminoids, which likely drove a moderate increase in upper soil carbon stocks. Soil respiration was consistently stimulated in watered plots, confirming the high sensitivity of soil microbes to moisture in dry tundra environments. As aboveground biomass and plant C uptake were not equivalently enhanced by watering, we suggest that belowground processes linked to root growth and/or microbial turnover are important in driving the C pool changes observed. Our results show evidence that increased summer rainfall can lead to greening and enhanced soil C storage in high arctic dry heaths, potentially providing moderate negative feedback to climate change.

Methods
Fieldwork was facilitated by the Zackenberg Ecological Research Operations station in July 2022. Plants and organic materials were collected and analysed in accordance with the non-exclusive licence G22-053 for utilization of Greenland genetic resources.

The experimental work was conducted in 2022 in the Zackenberg valley, NE Greenland (74°30’N, 21°00’ W).

The experiment consists of a fully factorial setup, with 48 plots of 0.5m × 0.5m dimensions in 6 blocks. To study the effects of increased summer precipitation and moderately increased nutrient availability, the full factorial block design includes 3 factors: Water (W), nitrogen (N) and phosphorus (P) additions, applied in all combinations within each block, hence with eight treatment combinations including controls in each block.

To analyse soil C pool changes >25 years after initiation of the experiment, four 10 cm-deep soil cores were taken from each of the 48 plots under the main vegetation types/species which were: Dryas, Salix, Kobresia and cryptogamic crust. The cores were separated into two depths:  0-5 (topsoil) and 5-10cm (mineral subsoil).

Aboveground vegetation cover in plots was estimated in late July 2022 using the point-intercept/point-framing method with 50 pin hits per plot. To determine plant %C, %N, 13C and 15N natural abundance, ten mature leaves of the three dominant plant species (Dryas, Salix and Kobresia) were collected in each plot.

All CO2 flux data was collected in summer 2022 using an EGM-4 portable Infrared Gas Analyser (PP Systems Unit 2, Hitchin, Hertfordshire, U.K). To measure the net ecosystem production (NEP), a transparent 8.43L plexiglass chamber was placed onto a 20 × 20 cm metal frame located in the centre of the plots (in place since 2004). To estimate Ecosystem Respiration (Reco), the chamber was covered with a black plastic cover and CO2 concentrations were measured as for NEP. To isolate the respiration component stemming from soil microbial sources (as well as a minor contribution from root respiration and faunal heterotrophs), also known as Rsoil,, we used a cylindrical closed system soil respiration chamber (SRC-1 – Probe type 8)

We measured CO2 flux in all 48 plots in 3 separate rounds within a two-week timespan (July 22nd -August 3rd, 2022) coinciding with peak growing season. In parallel to chamber flux measurements, we measured photosynthetic active radiation (PAR) using a PAR Meter (Sun System, U.S) and plot surface temperature with a Proscan 530 Dual focus infrared hand thermometer (Dostmann Electronic, Germany) at each plot prior to flux measurement. Soil volumetric water content and temperature were measured at 0-5 cm depth (TEROS-11 sensors, Meter Group AG, München, Germay) and logged every one hour (ZL6 Cellular Data Loggers, Meter Group AG, Müncen Germany) in all plots.