Associative learning of non-sugar nectar components: amino acids modify nectar preference in a hawkmoth

Abstract

The nearly ubiquitous presence of amino acids in the nectar of flowering plants has led to significant interest in the relevance of these compounds to pollinator behavior and physiology. A number of flower-visiting animals exhibit behavioral preferences for nectar solutions containing amino acids, but these preferences vary by species and are often context or condition dependent. Furthermore, the relative strength of these preferences and potential influence on the foraging behavior of flower-visiting animals remains unclear. Here, we used innate preference tests and associative learning paradigms to examine the nectar preferences of the flower-visiting hawkmoth Manduca sexta, in relation to both sugar and amino acid content. Manduca sexta exhibited a strong preference for higher sucrose concentrations, while the effect of amino acids on innate feeding preference was only marginally significant. However, with experience, moths were able to learn nectar composition and flower color associations and to forage preferentially (against innate color preference) for nectar with a realistic amino acid composition. Foraging moths responding to learned color cues of nectar amino acid content exhibited a behavioral preference comparable to that observed in response to a 5% difference in nectar sucrose concentration. These results demonstrate that experienced foragers may assess nectar amino acid content in addition to nectar sugar content and caloric value during nectar-foraging bouts.

Keywords: Manduca; Color; Fitness; Foraging behavior; Nutrition; Pollinator.

Fig. 1.

Foraging array. The foraging array consisted of a rotating disc with four artificial (blue/yellow) flowers positioned 20 cm apart. This was rotated 90 deg after every three visits from foraging moths to avoid positional learning.

Fig. 2.

Average reflectance spectra of the artificial paper flowers. n=6 for each color.

Fig. 3.

Innate preference indices of naive Manducasexta presented with experimental nectars of varying sucrose or amino acid content, compared with a standardized 25% sucrose control solution. Preference index (PI) values are shown on a scale of −1 to 1, with 0 indicating no preference. Negative values indicate a preference for the experimental solution, while positive values indicate a preference for the control solution. Experimental treatments containing only sucrose (numbered) are represented by white bars, and black bars represent experimental solutions containing sucrose plus amino acids (AA). Treatment abbreviations correspond to those in Table 2.

Fig. 4.

First-choice bias in first-choice test across 3 trial days. On day 1, prior to any experience, 86.3% of all first-choice visits were to yellow experimental flowers (χ²=38.479, d.f.=1, P<0.0001). This first-choice bias did not differ significantly in later trial days (P=0.26).

Fig. 5.

Color cues of nectar composition alter foraging behavior after learning. With experience, foraging moths were able to associate floral cues (i.e. color) with nectar composition and lessen the influence of innate biases. Rewarding treatments were: control (solid line), 20 versus 25 (dotted line) and 25 versus 25+AA (dashed line). Non-rewarding or aversive treatments were: 0 versus 25 (dot-dash line) and quinine (Q) versus 25 (long-dash line). Nectar and flower–color pairings are described in Table 3.