Taxonomic identification of pollen has historically been accomplished via light microscopy but requires specialized knowledge and reference collections, particularly when identification to lower taxonomic levels is necessary. Recently, next-generation sequencing technology has been used as a cost-effective alternative for identifying bee-collected pollen; however, this novel approach has not been tested on a spatially or temporally robust number of pollen samples. Here, we compare pollen identification results derived from light microscopy and DNA sequencing techniques with samples collected from honey bee colonies embedded within a gradient of intensive agricultural landscapes in the Northern Great Plains throughout the 2010–2011 growing seasons. We demonstrate that at all taxonomic levels, DNA sequencing was able to discern a greater number of taxa, and was particularly useful for the identification of infrequently detected species. Importantly, substantial phenological overlap did occur for commonly detected taxa using either technique, suggesting that DNA sequencing is an appropriate, and enhancing, substitutive technique for accurately capturing the breadth of bee-collected species of pollen present across agricultural landscapes. We also show that honey bees located in high and low intensity agricultural settings forage on dissimilar plants, though with overlap of the most abundantly collected pollen taxa. We highlight practical applications of utilizing sequencing technology, including addressing ecological issues surrounding land use, climate change, importance of taxa relative to abundance, and evaluating the impact of conservation program habitat enhancement efforts.

Identifying plant taxa that honey bees (Apis mellifera) forage upon is of great apicultural interest, but traditional methods are labor intensive and may lack resolution. Here we evaluate a high-throughput genetic barcoding approach to characterize trap-collected pollen from multiple North Dakota apiaries across multiple years. We used the Illumina MiSeq platform to generate sequence scaffolds from non-overlapping 300-bp paired-end sequencing reads of the ribosomal internal transcribed spacers (ITS). Full-length sequence scaffolds represented ~530 bp of ITS sequence after adapter trimming, drawn from the 5’ of ITS1 and the 3’ of ITS2, while skipping the uninformative 5.8S region. Operational taxonomic units (OTUs) were picked from scaffolds clustered at 97% identity, searched by BLAST against the nt database, and given taxonomic assignments using the paired-read lowest common ancestor approach. Taxonomic assignments and quantitative patterns were consistent with known plant distributions, phenology, and observational reports of pollen foraging, but revealed an unexpected contribution from non-crop graminoids and wetland plants. The mean number of plant species assignments per sample was 23.0 (+/- 5.5) and the mean species diversity (effective number of equally abundant species) was 3.3 (+/- 1.2). Bray-Curtis similarities showed good agreement among samples from the same apiary and sampling date. Rarefaction plots indicated that fewer than 50,000 reads are typically needed to characterize pollen samples of this complexity. Our results show that a pre-compiled, curated reference database is not essential for genus-level assignments, but species-level assignments are hindered by database gaps, reference length variation, and probable errors in the taxonomic assignment, requiring post-hoc evaluation. Although the effective per-sample yield achieved using custom MiSeq amplicon primers was less than the machine maximum, primarily due to lower “read2” quality, further protocol optimization and/or a modest reduction in multiplex scale should offset this difficulty. As small quantities of pollen are sufficient for amplification, our approach might be extendable to other questions or species for which large pollen samples are not available.

Colony monitoring devices used to track and assess the health status of honey bees are becoming more widely available and used by both beekeepers and researchers. These devices monitor parameters relevant to colony health at frequent intervals, often approximating real time. The fine-scale record of hive condition can be further related to static or dynamic features of the landscape, such as weather, climate, colony density, land use, pesticide use, vegetation class, and forage quality. In this study, we fit commercial honey bee colonies in two apiaries with pollen traps and digital scales to monitor floral resource use, pollen quality, and honey production. One apiary was situated in low-intensity agriculture; the other in high-intensity agriculture. Pollen traps were open for 72 h every two weeks while scales recorded weight every 15 min throughout the growing season. From collected pollen, we determined forage quantity per day, species identity using DNA sequencing, pesticide residues, amino acid content, and total protein content. From scales, we determined the accumulated hive weight change over the growing season, relating to honey production and final colony weight going into winter. Hive scales may also be used to identify the occurrence of environmental pollen and nectar dearth, and track phenological changes in plant communities. We provide comparisons of device-derived data between two apiaries over the growing season and discuss the potential for employing apiary monitoring devices to infer colony health in the context of divergent agricultural land use conditions.

Use of environmental DNA (eDNA) to assess distributions of aquatic and semi-aquatic macroorganisms is promising, but sampling schemes may need to be tailored to specific objectives. Given the potentially high variance in aquatic eDNA among replicate grab samples, compositing smaller water volumes collected over a period of time may be more effective for some applications. In this study, we compared eDNA profiles from composite water samples aggregated over three hours with grab water samples. Both sampling patterns were performed with identical autosamplers paired at two different sites in a headwater stream environment, augmented with exogenous fish eDNA from an upstream rearing facility. Samples were filtered through 0.8 μm cellulose nitrate filters and DNA was extracted with a cetyl trimethylammonium bromide procedure. Eukaryotic and bacterial community profiles were derived by amplicon sequencing of 12S ribosomal, 16S ribosomal, and cytochrome oxidase I loci. Operational taxa were assigned to genus with a lowest common ancestor approach for eukaryotes and to family with the RDP Classifier software for prokaryotes. Eukaryotic community profiles were more consistent with composite sampling than grab sampling. Downstream, rarefaction curves suggested faster taxon accumulation for composite samples, and estimated richness was higher for composite samples as a set than for grab samples. Upstream, composite sampling produced lower estimated richness than grab samples, but with overlapping standard errors. Furthermore, a bimodal pattern of richness as a function of sequence counts suggested the impact of clumped particles on upstream samples. Bacterial profiles were insensitive to sample method, consistent with the more even dispersion expected for bacteria compared with eukaryotic eDNA. Overall, samples composited over 3 h performed equal to or better than triplicate grab sampling for quantitative community metrics, despite the higher total sequencing effort provided to grab replicates. On the other hand, taxon-specific detection rates did not differ appreciably and the two methods gave similar estimates of the ratio of the common fish generaSalmoandCoregonusat each site. Unexpectedly,SalmoeDNA dropped out substantially faster thanCoregonuseDNA between the two sites regardless of sampling method, suggesting that differential settling affects the estimation of relative abundance. We identified bacterial patterns that were associated with eukaryotic diversity, suggesting potential roles as biomarkers of sample representativeness.

Mussels are useful indicator species of environmental stress and degradation, and the global decline in freshwater mussel diversity and abundance is of conservation concern. Elliptio complanata is a common freshwater mussel of eastern North America that can serve both as an indicator and as an experimental model for understanding mussel physiology and genetics. To support genetic components of these research goals, we assembled transcriptome contigs from Illumina paired-end reads. Despite efforts to collapse similar contigs, the final assembly was in excess of 136,000 contigs with an N50 of 982 bp. Even so, comparisons to the CEGMA database of conserved eukaryotic genes indicated that ∼20% of genes remain unrepresented. However, numerous candidate stress-response genes were present, and we identified lineage-specific patterns of diversification among molluscs for cytochrome P450 detoxification genes and two saccharide-modifying enzymes: 1,3 beta-galactosyltransferase and fucosyltransferase. Less than a quarter of contigs had protein-level similarity based on modest BLAST and Hmmer3 statistical thresholds. These results add comparative genomic resources for molluscs and suggest a wealth of novel proteins and noncoding transcripts.