Open AccessMonitoring of honey bee floral resources with pollen DNA metabarcoding as a complementary tool to vegetation surveys
Author(s) -
Milla Liz,
SchmidtLebuhn Alexander,
Bovill Jessica,
EncinasViso Francisco
Publication year - 2022
Publication title -
ecological solutions and evidence
Language(s) - English
Resource type - Journals
ISSN - 2688-8319
DOI - 10.1002/2688-8319.12120
Subject(s) - pollen , biology , taxon , ecology , biodiversity , vegetation (pathology) , pollen source , extinction (optical mineralogy) , pollination , botany , pollinator , medicine , pathology , paleontology
AbstractMonitoring biodiversity is a growing and pressing challenge, particularly as climate change threatens species with extinction and leads to widespread shifts in plant distribution and phenology. Tracking changes via ground vegetation surveys is costly and time‐consuming, hence monitoring of complex and heterogenous communities remains an ongoing challenge. Molecular DNA methods are rapidly being developed to provide fast and reproducible results for environmental monitoring, including diet and ecosystem assessments. Here, we used DNA metabarcoding of pollen foraged by European honey bees ( Apis mellifera ) to investigate their floral resource use in an urban reserve. We collected three different pollen samples from hives: individual bees, raw honey and pollen traps, and identified plants using two metabarcoding markers (ITS2 and trnL). We then compared the results to a ground vegetation survey of surrounding flowering taxa. Pollen DNA metabarcoding detected 74 taxa (48.6% identified to species) across all pollen sources, compared to 44 taxa recorded by the survey (93% identified to species). Within the metabarcoding results, we identified 25% of the genera and 9% of the species found during the survey, with three of the top 10 flowering genera represented. While honey was the most taxon‐rich pollen source (mean = 8.5, SD = 3.5), followed by honey bees (mean = 5.8, SD = 6.1) and pollen traps (mean = 4.2, SD = 1.7), combining the results of six individual bees could detect similar taxa numbers to honey, while 20 bees were required to detect as many taxa as the survey. We demonstrate how DNA metabarcoding of the pollen foraged by honey bees can detect more flowering taxa than traditional survey methods, and how different pollen sources and genetic markers affect the level of detection of plant taxa. The foraging choices of honey bees matched few species detected by the vegetation survey, therefore pollen metabarcoding is recommended as a complementary approach to ground surveys. Rigorous validation and stringent filtering of metabarcoding results were also required to exclude potential false positives. Altogether, this molecular approach can be used to augment vegetation surveys, while tracking the floral resources used by bees.