Data for: Parasitoids indicate major climate-induced shifts in Arctic communities

Climatic impacts are especially pronounced in the Arctic, which as a region is warming twice as fast as the rest of the globe. Here, we investigate how mean climatic conditions and rates of climatic change impact parasitoid insect communities in 16 localities across the Arctic. We focus on parasitoids in a wide-spread habitat, Dryas heathlands, and describe parasitoid community composition in terms of larval host use (i.e. parasitoid use of herbivorous Lepidoptera versus pollinating Diptera) and functional groups (i.e. parasitoids adhering to an idiobiont versus koinobiont lifestyle). Of the latter, we expect idiobionts to be generally associated with poorer tolerance to cold temperatures. To further test our findings, we assess whether similar climatic variables are associated with host abundances in a 22-year time series from Northeast Greenland. We find that sites which have experienced a temperature rise in summer while retaining cold winters to be dominated by parasitoids of Lepidoptera, with the pattern reversed among the parasitoids of Diptera. The rate of summer temperature rise is further associated with higher levels of herbivory, suggesting higher availability of lepidopteran hosts and changes in ecosystem functioning. We also detect a matching signal over time, as higher summer temperatures, coupled with cold early winter soils, are related to high herbivory by lepidopteran larvae, and to declines in the abundance of dipteran pollinators. Collectively, our results suggest that in parts of the warming Arctic, Dryas is being simultaneously exposed to increased herbivory and reduced pollination. Our findings point to potential drastic and rapid consequences of climate change on multitrophic-level community structure and on ecosystem functioning and highlight the value of collaborative systematic sampling effort.,1. Description of methods used for collection/generation of data: This dataset comprises of parasitoids caught in 2016 at 19 Arctic and Sub-Arctic localities during two consecutive six-day-long trapping periods aimed to take place during the flowering peak of the mountain avens (Dryas spp.). Each location had three to four trapping sites (A-D) in Dryas heath type habitats, each with ten 5cm by 4.5cm white sticky traps cut out from sticky board (Barrettine Environmental, UK [product no longer available]). Sticky traps were embedded in growths of Dryas spp. The parasitoids were subsequently picked of off the sticky traps, their whole DNA was extracted and half of their Cytochrome Oxidase I barcode region was amplified using Primers B-F and HCO. The processing of samples was done in three parts (Data1, Data2, Data2) with slightly different methodology. See the supplementary information of the recommended publication for more details. Datasets were sequenced at the Helsinki Functional Genomics Unit (FuGU) in two separate MiSeq v3 2x300bp runs (Data1 and Data2). Additionally, a set of samples from a specific site (Zackenberg) were sequenced as part of larger set at the FIMM Technology Centre in a HiSeq2500 2x250bp run (Data3). Additionally, Dryas flower count, flowering phenology and flower damage by insect herbivores was recorded at the start, after a week (day 6) and in the end (day 12). These counts were done in 3 to 5 1/4 square meter monitoring plots per trapping site. Microclimate was recorded at one trapping site per locality using Lascar EL-USB-2 tempeerature and air humidity loggers under a small white plastic dome at ~ 10 cm height. 2. Methods for processing the data: Initially, paired-end reads were merged and trimmed for quality using 32-bit usearch version 11 (Edgar 2010) with the command ‘fastq_mergepairs’. Primers were removed using software cutadapt version 1.14 (Martin 2011) with 15% mismatch rate. The reads were then collapsed into unique sequences (singletons removed) with command ‘fastx_uniques’. The subsequent clustering steps differed slightly for different data sets, due to the origin of the data (MiSeq vs. HiSeq2500), as follows. For Data1 and Data2, the newly-collapsed unique sequences were cleaned of chimeras using command ‘uchime_denovo’ and clustered into 96% OTUs (OTU = Operational Taxonomical Unit) using command ‘cluster_size’ using USEARCH. The choice of 96% clustering threshold was based on empirical optimization, considering both the rapid genetic divergence in CO1, as well as potential sequencing errors. For Data3, the unique sequences were denoised (i.e., chimeras were removed) and reads were clustered into ZOTUs (= ‘zero-radius OTU’) with command ‘unoise3’ using USEARCH version 11. These ZOTUs do not practically differ from traditional clustering of OTUs (which are based on pre-set percentage threshold), but according to Edgar and Flyvbjerg (2015), the UNOISE algorithm performs better for certain heterogenous data sets in (i) removing chimeras, (ii) PhiX sequences and (iii) Illumina artefacts. Then OTUs and ZOTUs were mapped back to the original trimmed reads with command ‘usearch_global’ (‘search_exact’ for Data3) to establish the total number of reads in each sample using 64-bit software vsearch (Rognes et al. 2016). Overall, we were able to map 25,673,920 reads (Data1: 4,261,291; Data2: 12,095,141; Data3: 9,317,488) to our original samples. These reads were subject to further filtering: from each sample, each OTU/ZOTU with less reads than 2% of the total reads in that sample were discarded, which also cleared most of the extraction and PCR negative controls. Finally, samples producing less than 37 reads (a threshold chosen by analysing the data as a whole) were removed from the subsequent analysis. The taxonomic assignations were initially done independently for each dataset (using identical criteria), but the final assignations were carried out using the whole, combined (Data1+Data2+Data3) dataset. The OTUs/ZOTUs were initially identified into genus-level using the RDP classifier with a very recently constructed COI-RDP database v3.2 (with 60% probability threshold for genus-level assignation) following Porter and Hajibabaei (2018). In cases where the database was clearly insufficient to reach a genus-level assignation, we used local BLAST against all the retrieved COI sequences in BOLD (Altschul et al. 1990; Ratnasingham and Hebert 2007) and chose the most probable match. Taxonomic information for remaining hits was retrieved manually from BOLD using BIN code (from earlier steps) or the actual OTU/ZOTU sequence. Finally, identifications were checked against our preliminary identification notes taken at the beginning of DNA extraction, and potentially false assignations (due to for example contamination in certain steps, or clear errors in the database) were either removed or assigned to the likely correct out/ZOTU. As the end result of all the bioinformatic steps, we arrived at a list of 460 parasitoid taxa (listed in OTU_info.csv Those dataset specific OTUs which were collapsed to one in the merging of the datasets are also listed in this file).,The included README_PanArcticParasitods.txt file contains detailed metadata on all included variables. There are some missing data e.g.: In the in situ microclimate measurements misising values are denoted with, NA In the parasitoid data, samples which failed to to produce parasitoid reads are marked with NA. In the Dryas data, two localities have their data data missing, and are not included in this table, where as one site Kobbejford had no Dryas and has counts of Loiseleuria procumbens as a phenological indicator in stead. This is marked in the datapoint notes. Concerning OTUs: Our OTUs are based on only a half of the CO1 barcode region and we used a quite wide clustering treshold of 96%. Also the clustering was done globally on a dataset containing closely related species from different continents. While this works well for the intended purpose of looking at functional group dominance at site level, the taxonomic resolution is inadequate for applications requiring true species level information. In such occasiton we encourage contacting the corresponding author. The DNA extracts are stored at the Department of Agricultural Sciences of the University of Helsinki, Finland.