B. Triwaks Bee Research Center, Department of Entomology, Robert H. Smith Faculty of Agriculture, Food and Environment, The Hebrew University of Jerusalem, Rehovot, Israel
Honey bees visit flowers to collect nectar, mostly as a source of carbohydrates, and pollen, which contains carbohydrates but in addition is rich in proteins, fats and minerals. As bees forage for food, they provide pollination services that are essential for maintaining our food security. Research in our lab has unravelled some of the intricacies of how the thousands (up to tens of thousands) of workers in a honey bee colony maintain nutritional homeostasis.
Balancing protein to carbohydrate ratio:
Honey bees evaporate the water from nectar, concentrating it into honey that is stored in the hive. A colony may store up to tens of kg of honey, which provides it energy for many weeks or months. Pollen, on the other hand, is stored in small quantities, which last a short time. Consequently, we found that the relative allocation of foraging effort for protein (P) or carbohydrate (C) by the colony is highly sensitive to its pollen stores 1. In control colonies we did not manipulate P and C stores, whereas in experimental colonies we removed honey stores (C-) and either removed (P-) or added (P+) pollen stores. The colony allocated 97% of foragers to collecting nectar in the P+ manipulation, relative to 54% in the P- manipulation and an intermediate 79% when not manipulated (Figure 1A). Based on the nutritional composition of the collected food, we could employ techniques of the geometric approach to nutrition to calculate the P:C target ratio of colonies in the different conditions. Colonies always bias their intake towards carbohydrates, but the bias is less (P:C = 1:14) when protein is lacking, P-, and very strong (P:C = 1:2,080) when protein is not lacking, P+ (Figure 1B).
Figure 1. A: The mean (SE) percentage of honey bee (HB) foragers visiting dishes with sucrose solution (nectar, N) or with pollen (P) when pollen stores are augmented (P+) or depleted (P-). Different letters in bars indicate statistically significant differences between groups (P<0.05). B: The mean (SE) amounts of protein and carbohydrate and the P:C ratio collected by colonies in the different treatments. * = statistically significant difference (P<0.05).
Balancing essential amino acids:
Pollens of different plants differ in their amino acid and fatty acid composition. Honey bee colonies may approach a balanced diet by collecting pollen from 5-8 different plant species concurrently 2. We tested whether a colony can actively bias its foraging towards sources that complement its specific amino acid deficiencies 3. Colonies were deprived of their pollen stores, then fed flours that were either poor or rich in specific essential amino acids (eAA). After a week of feeding, we provided foragers choice between dishes containing the same flour on which they had fed all week, another flour that was similar in eAA composition, or a complementary diet that would complement the eAA deficiencies of the flour they had been fed (Figure 2A). The experiment was balanced across colonies for which flours were used in the three nutritional roles. Colonies biased their foraging to the flour that complemented their colony’s eAA deficiencies (Figure 2B).
Figure 2. A: Top view photo of honey bees from one colony foraging from dishes with three types of flours, in a block design of eight replicates of the three dishes. B: Mean (SE) mg of collected diet per dish of flours that were the same, similar or complementary to the diet the colony had been fed previously. Different letters above bars indicate statistically significant differences between groups (P<0.05). Photo: H. Hendriksma.
Balancing essential fatty acids:
In a conceptually similar experiment as above, we fed a colony for several days only one type of pollen, that was biased towards high ratio of either one of the two essential fatty acids, omega-3 or omega-6, relative to the other one. But here we used observation hives, which have a glass panel through which we could record the recruitment dances of returning foragers. These dances consist of a bee running inside the hive on the comb in circles, reversing direction after each round. The rate at which the dancer performs these reversals is an indication of her subjective evaluation of the value of the food that she is advertising. After several days with one pollen, we presented for multiple 5-min periods a petri dish containing one of three pollens: the same pollen that the colony had been feeding on, another pollen that was similarly biased in omega-6:3 composition, or a pollen that was complementary in omega-6:3 composition. Dancers were individually marked, and we recorded their dance rate to the three pollens. The “same” pollen elicited the least excited dances, whereas the “complementary” pollen elicited the fastest dances4. Thus, foragers can assess the colony’s specific needs and bias their recruitment towards resources that best fulfil these needs.
Figure 3. A,B: Marked foragers collecting pollen from dish, and on the dance floor. B: Mean (SE) dance rate to pollen that is the same, similar or complementary to the pollen the colony had been fed previously. Numbers at bottom of bars are dance sample sizes. Different letters above bars indicate statistically significant differences between groups (P<0.05). LSMeans are those predicted by the full statistical model. Photos: S. Zarchin, K. Oxman.
Cognitive impairment of unbalanced omega-6:3 ratio:
A high omega-6:3 ratio in the diet of mammals (including humans) is associated with cognitive decline. We tested whether honey bees would be similarly affected. Colonies were fed either artificial diets skewed towards high omega-6:3 ratio (by enriching the diets with corn or sesame oil) or with a more balanced omega-6:3 ratio (oil mixture that included sage and flax oils). Another balanced diet consisted of a mixture of pollens. We then tested the associative learning performance of bees that developed on these diets using the proboscis extension response (PER) conditioning assay (Figure 4A). Each bee experienced six trials in which an odor (conditioned stimulus, CS+) was associated with feeding a drop of sucrose solution, and six trials in which a different odor (CS-) was associated with exposure to salt solution, which is aversive to bees. The order of CS+ and CS- trials was according to a pseudorandom sequence. The percentage of bees that responded to the CS+ odor, and prior to receiving the sucrose reward, over subsequent trials represented a learning curve. Bees fed a balanced omega-6:3 diet achieved about 95% learning, whereas learning performance of those fed an omega-6 biased diet was greatly impaired, reaching only about 40% (Figure 4B).5,6
Figure 4. A
: A honey bee worker in conditioning of the proboscis extension response (PER), is fed a drop of sucrose solution after exposure to the conditioned odor. B: Learning curves of bees that developed on a high (corn and sesame) or low (oil mixture and pollen) omega-6:3 ratio diet, showing about 95% PER to the rewarded odor (CS+) in the latter group vs only about 40% in the former (P<0.0001); response to the punished odor (CS-) was very low in all groups. Photo: Y. Arien.
I greatly acknowledge the dedicated and hard work of my postdoc, Harmen Hendriksma, students, Shlomi Zarchin and Yael Arien, technician, Haim Kalev, collaborator, Arnon Dag, and various additional assistants.