A learning assay for the color discrimination skills of the Amazon molly

In an ever-changing world, animals must constantly adjust their behavior and learn new associations. These can be learned from trial-and-error or observing conspecifics and/or other species. In this study, the color discrimination and learning abilities of the clonal Amazon molly (Poecilia formosa) were assessed in a simple binary choice assay. Two groups of genetically identical individuals were tested on alternate days and were required to learn to form a connection between a certain color and a hidden food reward inside the colored compartment (Red or Blue). The experimental results indicated that the fish potentially learned to make associations with color and food as they appeared to improve in accuracy over time. The average proportion of trials with the blue colored cup entered first increased from 25% (on the trial day 1) to 54% (day 5), with a correlation of r = 0.962 (p = 0.009). This finding however suggests a day 5 accuracy of no better than random chance. We also found that the fish who were assigned the focal color red performed significantly better compared to those assigned to blue (89% and 44% as the accuracy, respectively; 2-sample t test p < 0.001). Overall, the results suggested that this assay was unsuccessful in providing evidence for the color discriminating abilities - and hence, of the changes in learning abilities - of the clonal Amazon molly. The unsuitability of this assay may be due to its inability to discount a preference for the red color; being similar to the color of the diet of the Amazon molly.