A century of wild bee sampling: historical data and neural network analysis reveal ecological traits associated with species loss

Authors: Graham, Kelsey; GLAUM, PAUL - University Of California, Davis; HARTERT, JOSEPH - University Of Michigan; GIBBS, JASON - University Of Manitoba; TUCKER, ERIKA - Milwaukee Public Museum; ISAACS, RUFUS - Michigan State University; VALDOVINOS, FERNANDA - University Of California, Davis

Submitted to: Proceedings of the Royal Society of London B
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 7/12/2024
Publication Date: 8/14/2024
Citation: Graham, K.K., Glaum, P., Hartert, J., Gibbs, J., Tucker, E., Isaacs, R., Valdovinos, F. 2024. A century of wild bee sampling: historical data and neural network analysis reveal ecological traits associated with species loss. Ecology. https://doi.org/10.1098/rspb.2023.2837.
DOI: https://doi.org/10.1098/rspb.2023.2837

Interpretive Summary: Wild bees are critical for pollination of plant communities. There has been evidence of wild bee decline worldwide, yet there have been few projects in the United States that analyze changes in the bee community over a long time period. We compared changes in the wild bee community from 1921 to 2018 at a nature preserve in southern Michigan, USA. Across the near century of records, there were 208 bee species detected at the preserve, with year to year changes in the number of species detected. During an intensive survey in 1972 and 1973 by F.C. Evans, 135 bee species were detected. In the most recent intensive surveys conducted in 2017 and 2018, there were 90 species recorded. We found that measurements of the bee community diversity (diversity, species richness, and evenness) were all lower in recent samples. There was also evidence of declining bee abundance. Of 22 bee species collected in both periods and with at least 30 records, 64% of these bees declined in abundance. We then compared bee species traits (nesting habits, pollen preferences, geographical spread, etc.) using neural networks, revealing that ground nesting bees with specialized pollen preferences(only collect pollen from a few plant species) and cleptoparasitic bees were less likely to be found in recent surveys, whereas cavity nesting bees with a wider breadth of pollen preferences were more likely to have persisted. Larger bee body size was also associated with increased probability of local extirpation, though only for cavity nesting species with wide dietary breadths. Bees in the contemporary samples also had a more southerly overall distribution compared to the historic one. This study provides evidence for significant bee community shifts in the preserve since the 1970s, and identifies traits associated with persistence or local extirpation. We also discuss the challenges with using museum data for detecting bee community change over time.

Technical Abstract: We analyzed the wild bee community sampled from 1921 to 2018 at a nature preserve in southern Michigan, USA to determine whether there were significant shifts in the bee community over this timespan. Across the near century of records, there were 208 species detected at the preserve, with year to year variations in the number of species detected. During an intensive survey in 1972 and 1973 by F.C. Evans, 135 bee species were detected. In the most recent intensive surveys conducted in 2017 and 2018, there were 90 species recorded. Only 58 species were recorded in both sampling periods, indicating a significant shift in the bee community. We found that the bee community diversity, species richness, and evenness were all lower in recent samples. There was also evidence of declining abundance in many of the more common species. Of 22 bee species collected in both periods and with at least 30 records, 64% of these bees exhibited a decline in relative abundance of over 30%, and five of these species had more than a 95% reduction in captures. We also conducted traits analyses using neural networks, revealing that oligolectic ground nesting bees and clepto-parasitic bees were less likely to be found in recent samples whereas polylectic cavity nesting bees were more likely to have persisted. Larger body size was also associated with increased probability of local extirpation, though only for polylectic cavity nesting species. Larger phenological range was associated with increased chances of persistence for polylectic cavity nesting species and polylectic ground nesting species, while it was associated with increased rates of extirpation for oligolectic ground nesting species. Bees in the contemporary samples also had a more southerly overall distribution compared to the historic one. This study provides evidence for significant bee community shifts in the preserve since the 1970s, and identifies traits associated with persistence or local extirpation. We also discuss the challenges with using museum data for detecting bee community change over time.