Data from: Predator density outweighs experimental warming effects on short-term carbon and nitrogen loss from arctic shrub litter
Data files
Oct 09, 2025 version files 13.59 KB
-
README.md
4.77 KB
-
Sagi_BetulaLitterData_Oikos.csv
8.82 KB
Abstract
Rapid climate change in the Arctic is altering biological communities and their subsequent effects on ecosystem functioning. For example, warming-induced shrub expansion accelerates biogeochemical cycles in part by increasing high-quality litter inputs. Likewise, warming may enable higher densities of wolf spiders, which are dominant invertebrate predators whose activities indirectly alter plant litter decomposition rates. Although shrubs and wolf spiders are responding to climate change simultaneously, it is unclear how more shrub litter and more spiders together will influence elemental cycling in Arctic ecosystems. To test how warming could influence these processes, we used a fully factorial mesocosm experiment to quantify effects of wolf spiders on litter decomposition of an expanding species of dwarf deciduous shrub (Betula nana) under ambient and warmed conditions. We found higher densities of wolf spiders were consistently associated with more litter mass loss and more C and N release, regardless of warming treatment, indicating biotic interactions may be a stronger driver of short-term B. nana litter decomposition than warming when wolf spiders are present. Our findings suggest the combined effects of warming-induced shifts in plant and arthropod communities may further accelerate C and N cycling, which could cause positive feedbacks on Arctic shrub expansion.
Dataset DOI: 10.5061/dryad.9p8cz8www
This dataset was produced from a field mesocosm experiment near Toolik Field Station (68°38′N and 149°43′W, elevation 760 m) on the North Slope of Alaska during June-July 2012. We investigated the effects of wolf spider densities and experimental warming (using open-topped chambers) on mass, C and N loss from litter of Betula nana, which is an expansive shrub in the Arctic.
Five blocks were distributed in a homogenous area of moist acidic tundra, with each block containing a plot for each of the six experimental treatments, which were all possible combinations of the temperature treatment (ambient/high) and the spider density treatment (low/control/high).
We deployed three replicate litter bags at evenly spaced locations within each plot.
Each row in the dataset represents an individual litter bag.
Mesocosms were assigned one of three wolf spider density treatments: low density, control density, and high density. Control densities simulated natural wolf spider densities (approximately 1.1 spider per m2) based on field surveys conducted in an area adjacent to our experimental plots. For low density plots, we removed all wolf spiders prior to the start of the experiment and maintained the treatment by periodically removing stray wolf spiders that had entered the plots throughout the course of the experiment. High density treatments received enough wolf spiders to double the density of control plots at the beginning of the experiment and subsequently received additional wolf spiders throughout the summer if densities declined. We used visual surveys at the end of the field season to confirm that plot-level wolf spider densities varied according to their pre-assigned treatments (ANOVA: F2,27 = 21.85, P = < 0.0001), with average densities in high-density plots being 3.3 ± 0.47 wolf spiders, control plots 1.8 ± 0.20 wolf spiders, and low-density (removal) plots with 0.3 ± 0.213 wolf spiders each. In the year of our experiment, the wolf spider community in this habitat type around Toolik Lake was dominated by a single species (>95%), Pardosa lapponica.
Half of the mesocosms were experimentally warmed using ITEX open-topped warming chambers (OTCs), which are a common experimental method for passive warming in tundra ecosystems. This approach typically increases mean daytime air temperature by 1–2 °C and surface temperature by 1.27 °C. OTCs were positioned over the mesocosm plots during the summers of 2011 and 2012 immediately after snow melt. In both years, the experimental warming treatment was in place from approximately mid-June to early August.
Description of the data and file structure
The dataset contains one csv file:
Sagi_BetulaLitterData_Oikos.csv
Variables
block: the spatial block in which the plot was located, within which the litter bag was deployed.
plot: the plot within which the litter bag was deployed.
year: this particular dataset is only from 2012.
location: in this dataset there are only litter bags that were deployed on the soil surface.
weeks incubation: 6 for all litter bags in this dataset.
wolf spider density treatment: low/control/high
temperature treatment: ambient/high
average soil moisture: we measured soil moisture at each litter bag location at the beginning, middle, and end of the field season using a HydroSense portable soil moisture probe. The average of these measurements is reported for each litter bag.
litter percent moisture [%]: At the end of experiment, we estimated litter moisture content by weighing, drying at 60 °C for 72 hours and reweighing.
percent N remaining [%]: Initial and final N content was measured using an elemental analyser feeding a continuous flow mass spectrometer system.
percent C remaining [%]: Initial and final C content was measured using an elemental analyser feeding a continuous flow mass spectrometer system.
The % C and N contained within each sample were converted to mass based on the pre- or post-decomposition (initial and final respectively) sample weight and then % of C, N, and total mass remaining were calculated as follows:
% 𝑟𝑒𝑚𝑎𝑖𝑛𝑖𝑛𝑔 = ([%𝑤/𝑤]𝑓 ∗ [𝑚𝑎𝑠𝑠]𝑓) / ([%𝑤/𝑤]𝑖 ∗ [𝑚𝑎𝑠𝑠]𝑖) ∗ 100
where %𝑤/𝑤 represent mass percent of the element in a sample, whereas 𝑓 and 𝑖 represent final and initial values.
PercentMassRemaining [%]: Percent mass remaining was calculated as: (final dry litter mass / initial litter mass) * 100
Code/software
Any software that can open csv files