Transcriptomics, the quantification of gene expression, provides a versatile tool for ecological monitoring. Here, we show that through genome-guided profiling of transcripts mapping to 33,042 gene models, expression differences can be discerned among multi-year and seasonal leaf samples collected from American beech trees at two latitudinally separated sites. Despite a bottleneck imposed due to large-scale post-Columbian deforestation, the SNP-based population genetic background analysis has yielded sufficient variation to account for differences between populations and among individuals. Our time series of expression analyses during spring-summer and summer-fall transitions for two consecutive years involved 4,197 differentially expressed protein-coding genes. Using Populus orthologs of the differentially expressed genes, we reconstructed a protein-protein interactome as a representation of the leaf physiological states of trees during the seasonal transitions. Gene set enrichment analysis revealed GO terms that highlight molecular functions and biological processes possibly influenced by abiotic forcings such as recovery from drought and response to excess precipitation. Further, based on 324 co-regulated transcripts, we focused on a subset of terms that could be putatively attributed to phenological shifts due to late spring. Our conservative results indicate that extended transcriptome-based monitoring of forests can capture ranges of responses arising from other factors including air quality, chronic disease as well as herbivore outbreaks that require activation and/or downregulation of genes collectively tuning reaction norms needed for the survival of long-living trees such as the American beech.

American beech leaves have been felled from the canopy through Spring, Summer and Fall between 2017 and 2018. For subcanopy trees we used a 12m pruner pole. Sample collections were done in the morning to avoid expression changes throughout the day. For each sample, 10 leaf discs were punched into 2ml extraction vials containing 3 steel balls of 3mm in diameter, and flash frozen in liquid nitrogen in the field (SI Fig. S2). Frozen samples were then transferred to a -80 freezer at the end of the day. RNA extractions were carried out using E.Z.N.A Plant RNA Kit by Omega Bio-Tek (Norcross, GA, USA). Quality metrics for extracted RNA were determined using the Agilent Bioanalyzer 2100 instrument (Agilent Technologies, Santa Clara, CA) and samples containing genomic DNA contaminants were treated with DNAse I (ThermoFisher Scientific, Waltham, MA). Illumina TruSeq libraries were prepared for each sample RNA (Illumina Inc., San Diego, CA, USA). Libraries were indexed and pooled to be sequenced on the Illumina NovaSeq 6000 platform as 150 nucleotide long paired end reads targeting ~14 million total reads per sample (Illumina Inc., San Diego, CA, USA). Sequences are uploaded to the NCBI Sequence Read Archive (SRA) with the bioproject number PRJNA630305.