1205

Original Paper: Bartomeus, I., J. S. Ascher, D. Wagner, B. N. Danforth, S. Colla, S. Kornbluth et al. 2011. Climate-associated phenological advances in bee pollinators and bee-pollinated plants. Proceedings of the National Academy of Sciences USA 108: 20645−20649.

Rising global temperatures due to climate change affect the timing of natural events, including insect pollination and other species interactions. As temperatures rise, these events happen earlier. Not all species react to these changes in similar ways, and this can result in timing mismatches. For example, if the timing of insect emergence varies too much from that of flowering, pollination may not occur.

Scientists looked at 3,447 museum specimens of native bee species dating back to the 1870s. They chose 10 bee species that are known to range across North America, emerge in early spring, and pollinate both crops and wild plants. The scientists then compared the predicted pollination date for individual bees with flowering times of bee-pollinated spring plants measured in other studies. Dates of bee pollination activity and flowering times were correlated with temperature data in the same geographic areas.

Figure A shows scatter plots of the distribution of predicted pollination dates for individual bees between 1870 and 2010 (top) and the mean April air temperature between 1900 and 2010 (bottom). Predicted pollination date is the number of days between January 1 and the bee-collection date, a time in which bees were assumed to be engaging in pollination activity. Black lines indicate the trends for the entire time period, and red lines show the trends between 1970 and 2010. Statistical analysis revealed a strong positive correlation between predicted pollination dates and mean April air temperature. Between 1880 and 2010, bees advanced their pollination activity by 10.4 days; most of the advance (7.2 days, or 69 percent) has occurred since 1970.

Figure B shows comparisons of the change in days for bee pollination time and plant flowering time for two time intervals. The change in slope between the two dates (bees at left, plants at right) indicates the degree of difference between bee pollination time and plant flowering time for both time intervals. The interval indicates a mismatch of 4−6 days over the 118-year span of the longer study (1885–2003). For the more recent study (1971−1999), when mean April air temperature changed the most, the mismatch was 5 days over just 30 years.

Questions

Question 1

Pollination is a positive interaction, or facilitation. Specifically, it is a mutualism, meaning both partners benefit. Bees obtain food (pollen and nectar) from plants. Plants rely on bees for pollination, which is necessary for successful reproduction. For a successful pollination interaction, bees must emerge at the same time that plants flower.

Question 2

Over time, the trend in bee pollination date is slightly downward, indicating that the bees were pollinating slightly earlier in the year. The mean April temperature showed an upward trend, indicating that temperatures rose over the time period. The opposing directions of the trends indicate that as temperatures rose, bees were pollinating earlier in the year. The steeper slope after 1970 on both graphs suggests that temperature has increased faster since 1970. Statistical analysis bears this out; 69 percent of the bees’ advance in pollination activity has occurred since 1970.

Question 3

The recent study (1971-1999) shows plant flowering advancing faster than bee pollination dates, and thus best supports the hypothesis that climate warming is affecting plant-pollinator interactions by creating mismatches in the timing of these events.

Question 4

Specialist plant-pollinator interactions are more likely to be affected by timing mismatches caused by rising temperatures. Generalist species (such as those in these studies) can rely on many species for food or pollination interactions; if one or several species pairs show timing mismatches, other species will likely be present to participate in the interaction. Specialist pairs, which depend on only one or a few species, are far more susceptible to the effects of timing mismatches. If timing mismatches occur, there may be no species available to interact with.

Question 5

There are many possible designs, but more precise information would require analysis of specific plant-pollinator pairs, rather than the generalized groups in the current studies. The study would require seasonal measurements of plant flowering dates and pollinator emergence and pollination activity dates (as opposed to collection dates) for several years, correlated directly with climate data. Ideally studies should focus on pairs for which background information is already available on past emergence and flowering dates. Good candidates would include pollinators of important crop plants, such as apple trees or tomatoes.