Data for: Carbon flow to generalist predators shifts between plant- and fungal-derived resources in highly seasonal ecosystems
Data files
Sep 16, 2026 version files 58.83 KB
-
README.md
5.35 KB
-
seasonal_data.csv
19.75 KB
-
site_year_data.csv
16.04 KB
-
Spider_Data.xlsx
17.69 KB
Abstract
Animals influence biogeochemical processes by consuming other organisms, transferring energy from basal resources up the food chain. Energy is routed through plant, fungal and bacterial channels, and the relative importance of these pathways can shift across space and time. Variation in these pathways may be particularly pronounced in highly seasonal environments where resource availability fluctuates throughout the year. Spatiotemporal variation in the transfer of plant- versus microbial-derived energy to higher trophic levels is poorly understood but can affect bottom-up and top-down processes. For example, seasonal shifts in herbivore versus microbivore availability (bottom-up) may alter prey choice by generalist-feeding predators and lead to changes in top-down predator control on primary production and decomposition. We tested how spatiotemporal variation in primary production influences the use of different energy channels by a widely distributed generalist invertebrate predator in the Arctic tundra. We sampled wolf spiders across four sites spanning a productivity gradient in the Alaskan Arctic during peak summer and on a weekly basis at one focal site throughout the growing season. Using essential amino acid (EAA) carbon isotope (δ¹³C) fingerprinting, we traced the flow of plant-, bacterial- and fungal-derived EAAs to spiders, and quantified the use of plant- versus microbe-based energy channels between sites and across summer. Wolf spiders were associated with more plant- and less fungal-derived carbon assimilation when larger or collected from sites with higher normalized difference vegetation index (NDVI). Across the growing season, spiders shifted from primarily assimilating fungal-based resources in early summer to plant-derived resources in late summer, aligning with seasonal increases in NDVI. These patterns demonstrate spatiotemporal variation in carbon flow through different energy channels and suggest subsequent top-down predator effects may also vary with basal resource availability. Our findings also suggest wolf spider body size may predict whether predation indirectly influences primary production or decomposition. As climate change alters seasonality, resource dynamics, and invertebrate body sizes, these insights provide a basis for predicting spatiotemporal variation in predator impacts on ecosystem processes. Finally, our work demonstrates the utility of compound-specific stable isotope analysis for integrating ecological and biogeochemical theory through the conceptual lens of zoogeochemistry.
Dataset DOI: 10.5061/dryad.g1jwstr84
Description of the data and file structure
We used CSIA to quantify how plant- and microbial-derived energy contribute to wolf spider (family Lycosidae) diets across spatial, seasonal, and interannual gradients in the Alaskan Arctic. First, we investigated how reliance on plant- versus microbial-derived energy varies with primary productivity by analyzing EAA δ¹³C values of individual wolf spiders collected from four sites spanning a latitudinal and productivity gradient as measured by normalized difference vegetation index (NDVI). Second, we examined how reliance on these energy channels varies across the summer season and in response to short-term increases in primary productivity by estimating weekly resource use at a focal site from early to peak summer. We also tested whether wolf spider body size explained variation in the relative contribution of plant- and microbially-derived energy channels.
Files and variables
File: Spider_Data.xlsx
Sheet: Site.Year
Description: includes consumers mass and LDA scores along with NDVI scores for the site and year they were collected in.
Variables:
- ID - The spider specimen ID.
- Site - Where specimen was collected.
- Year - When specimen was collected.
- Spider.dry.mass.mg - Specimen dry mass in mg, used as a continuous variable in MixSIAR models
- Mass_quartile - Division of spider mass distribution to quartiles (Q1,Q2,Q3,Q4), used as a factor in MixSIAR models.
- NDVI - Normalized Difference Vegetation Index. For each site, we calculated daily NDVI from Red and NIR bands estimated by the MODIS Surface Reflectance product at a 500 m spatial resolution (MOD09GA version 6.1) (Vermote & Wolfe, 2021). NDVI was then integrated from the approximate onset of the growing season (June 1st) until the date of spider collection for each year (2013 or 2021), excluding values smaller than 0.05 to avoid noise and artefacts.
- LD1 - The first linear discriminant score from a Linear Discriminant Analysis of essential amino acids 13C data.
- LD2 - The second linear discriminant score from a Linear Discriminant Analysis of essential amino acids 13C data.
Sheet: Seasonal
Description: includes consumers mass and LDA scores along with NDVI scores for the week they were collected in.
Variables:
- ID - The spider specimen ID.
- week - Ordinal, when specimen was collected.
- Date - When specimen was collected. Date format M/DD/YYYY.
- Spider.dry.mass.mg - Specimen dry mass in mg, used as a continuous variable in MixSIAR models.
- Weight_quartile - Division of spider mass distribution to quartiles (Q1,Q2,Q3,Q4).
- Size - Division of spider mass distribution to halves (Small,Large), used as a factor in MixSIAR models.
- NDVI (in Seasonal sheet) - Normalized Difference Vegetation Index. NDVI was integrated across seven days preceding spider collection.
- LD1 - The first linear discriminant score from a Linear Discriminant Analysis of essential amino acids 13C data.
- LD2 - The second linear discriminant score from a Linear Discriminant Analysis of essential amino acids 13C data.
File: site_year_data.csv
Description: includes interannual multi-site sources and consumers essential amino acids 13C data for linear discriminant analysis.
Variables:
- ID - specimen ID
- group1 - source type (bacteria, fungi, plant) for sources, site for consumers
- group2 - source/consumer
- site - where specimen was collected
- Al other variables are mean δ13C and standard deviation for each essential amino acid.
File: seasonal_data.csv
Description: includes seasonal within-site sources and consumers essential amino acids 13C data for linear discriminant analysis.
Variables:
- ID - specimen ID
- site - where specimen was collected
- date - when specimen was collected. Relevant for consumer data, n/a (non-applicable) for source data. Date format M/DD/YYYY.
- group1 - source type for sources, week of collection (ordinal - 1 to 6) for consumers
- group2 - source/consumer
- All other variables are mean δ13C and standard deviation for each essential amino acid.
Code/software
Data can be accessed by any software that is compatible with csv and xlsx files.
We used R software:
MASS package for LDA analysis
vegan package for permutational multivariate analysis of variance
MixSIAR package for Bayesian mixing models
Access information
Other publicly accessible locations of the data:
- N/A
Data was derived from the following sources:
-
Larsen, T., Taylor, D.L., Leigh, M.B. and O'Brien, D.M. (2009), Stable isotope fingerprinting: a novel method for identifying plant, fungal, or bacterial origins of amino acids. Ecology, 90: 3526-3535. https://doi.org/10.1890/08-1695.1
(Appendix B)
-
Manlick, P.J., Perryman, N.L., Koltz, A.M. et al. Climate warming restructures food webs and carbon flow in high-latitude ecosystems. Nat. Clim. Chang. 14, 184–189 (2024). https://doi.org/10.1038/s41558-023-01893-0
