Laboratory sample processing
We used morphological traits to identify insects to family. After identification, we dried insect samples at 60 ºC for 7 days before collecting total dried insect biomass for each plot. We sent soil microbial samples (collected in 2019 (WY) and 2020 (MT)) to Oregon State University’s Center for Genome Research and Biocomputing for extraction, amplification, cleanup, and sequencing to generate 250-bp paired-end reads. We then merged the forward and reverse reads using FLASH (Magoč and Salzberg 2011) and the merged reads were quality-controlled using the DADA2 pipeline (Callahan et al. 2016) in Qiime2 (Hall and Beiko 2018). This removed chimeras, and the data were filtered to further remove mitochondria, chloroplast, and unassigned sequences. We then removed any amplicon sequence variants with fewer than 10 reads and clustered sequences into operational taxonomic units (OTUs) with a 97% similarity threshold using the Silva database (Pruesse et al. 2007) for reference. We continued to follow the Qiime2 pipeline at a sampling depth of 30,000, which only excluded one plot with less than 8,000 reads.