Fig. 2.

The bimodal feeding pattern on adenine nucleotides in Aedes aegypti (a) and Anopheles gambiae (b) showing a dose-dependent increase in the volumes of AMP, ADP, and ATP imbibed. Different lowercase letters above bars indicate significant differences within ligands, with the single and double primes referring specifically to ADP (') and ATP (''), respectively, as determined by pairwise post-hoc tests. Asterisks indicate a significant difference between ligands, determined by two-way analysis of variance at **P = 0.0010 and ***P < 0.001. For Aedes aegypti, the number of replicates for each diet was N = 9, with each replicate consisting of n = 90 Ae. aegypti (row effect: F_{(11, 1184)} = 38.90; P < 0.001; column effect: F_{(2, 1184)} = 142.1, P < 0.001); for An. gambiae, the number of replicates for each diet was N = 5, with each replicate consisting of n = 50 An. gambiae (row effect: F_{(7, 735)} = 7.758; P < 0.001; column effect: F_{(2, 735)} = 8.223, P < 0.001). A main effects model was used due to the different dynamic range of concentrations tested for each ligand. ADP, Adenosine diphosphate; AMP, adenosine monophosphate; ATP, adenosine triphosphate; ns, non-significant