Maia, C. M. and Volpato, G. L. (2019), What to choose when the best preference is not available: does the Nile tilapia follow a linear sequence of preferences? Journal of Zoology, vol. 307, pp. 274-281. https://doi.org/10.1111/jzo.12646
Our interest for studying fish preferences came to us because of welfare concerns, where it is supposed that fish have better welfare when they are in conditions they freely prefer. This raised the need for differentiation between preference and spurious choice. In previous studies we could differentiate more consistent responses – preferences – from just spurious choices – non-preferences – of fish. More studies were then carried out to better understand fish preferences. Firstly, we detected an individual and more consistent response over time. Since fish prefer some resources over others, mainly investigated as background colors in our studies, an intriguing emerging question was to know what they would select consistently when what they prefer is not available. Do they have a linear scale of preferences, so that the second preference replaces the absence of the first one? To detect and choose the best enrichment resources for animals to improve their welfare conditions, this is an important question. After an individual’s preferred resources are ranked, should we provide the animal with the second ranked resource to replace the first one?
We chose to investigate this question in the Nile tilapia, Oreochromis niloticus (Figure 1), because this is like our pet lab animal, and also because we had already detected some resource preferences in this species. Moreover, some items of preference are individual responses for Nile tilapia that vary inside a group, as seen in previous studies. Despite investigating just one fish species in this study, our conclusions can be easily understood for other objects and animals. If preferences do not follow a rigid sequence, then if the first preference is missing, another one needs to be available to maintain welfare conditions, however finding the new resource from a previous list of ranked resources is not valid.
In this context, we realized that by blocking the most preferred resource for these fish we could see what their responses are towards those resources that are left. From previous studies, we knew that Nile tilapia is a species that recognizes, chooses or prefers some environmental colors. We then developed a 4-compartment apparatus to test preferences for background colors (Figure 2). In the first 10-day period we tested individual fish to detect what colors they prefer and to rank their preference and non-preference responses. Then, we skipped one day and started another 10-day period in which we blocked access to the most preferred color of each individual (i.e. the entrance to the colored compartment was visually and physically blocked; Figure 2). During the two test periods, we recorded daily the fish positions once every 30 seconds for one hour. From these data, we used a Preference Index (PI), (Maia & Volpato, 2016, Scientific Reports, 6: 28328), which indicates by positive values the preferred colors and by negative values the non-preferred colors (spurious choices). Furthermore, the intensity of the preference or non-preference was indicated by the value of each PI.
In the first test period, all the 14 fish preferred at least one color. Preference for only one color signifies a more decided preference response than when the fish preferred two or more colors. The intensity of the PI ranked the whole set of preferred and non-preferred items for each fish. Based on these results, we found that in the second period of the test the new most preferred color was not always the previous second ranked color in the first period of the test. Although this was an exciting result, we asked whether such pattern of response could be attributed to the intensity of the preference response for the first or the second ranked color in the first period of the test. But this was not the case. Finally, we tested whether the intensity of the new preference in the second period of the test, that is, when the previous most preferred item was blocked, was higher when the fish selected the second ranked color from the first period of the test as the new most preferred item. This was also not the case, meaning that there was no effect of preference rank on the preference responses of the fish after the most preferred resource was removed. We then assumed that such change in the preference responses might be contributed to a change in the environmental context (from four to three colors as choice options).
These results highlight the importance to consider animal preferences in specific environments, instead of considering the resource preferences throughout different environmental contexts. This is even relevant when considering ornamental aquarium contexts, for example, because although the Nile tilapia is not considered an ornamental species, it is not expected that our findings are restricted to this species. Therefore, if you have an ornamental aquarium, for example, with several fish (Figure 3), pay attention that each one might prefer (and not just being displaced by dominant animals) different resources. Moreover, what they prefer after you have changed the environment (by adding new enrichments or species) might be reconsidered to ensure better welfare conditions.
Caroline Marques Maia