“Evolution of Sociability by Artificial Selection”, Andrew M. Scott, Ian Dworkin, Reuven Dukas2021-10-04 (, , ; similar)⁠:

There has been extensive research on the ecology and evolution of social life in animals that live in groups. Less attention, however, has been devoted to apparently solitary species, even though recent research indicates that they also possess complex social behaviors.

To address this knowledge gap, we artificially selected on sociability, defined as the tendency to engage in nonaggressive activities with others, in fruit flies. Our goal was to quantify the factors that determine the level of sociability and the traits correlated with this feature.

After 25 generations of selection, the high-sociability lineages showed sociability scores about 50% higher than did the low-sociability lineages. Experiments using the evolved lineages indicated that there were no differences in mating success between flies from the low and high lineages. Both males and females from the low lineages, however, were more aggressive than males and females from the high lineages. Finally, the evolved lineages maintained their sociability scores after 10 generations of relaxed selection, suggesting no costs to maintaining low and high sociability, at least under our settings.

Sociability is a complex trait, which we currently assess through genomic work on the evolved lineages.

[Keywords: aggression, artificial selection, Drosophila melanogaster, mate choice, sociability, social behavior]

…Field and laboratory studies indicate that both larval and adult fruit flies show statistically-significant sociability, as they prefer to group together at food patches (Durisko et al 2014; Anderson et al 2016; Scott et al 2018; Dukas2020). In the adults, the broad sense heritability of sociability is about 0.22 (Scott et al 2018). The heritable variation in sociability opens up exciting opportunities for assessing the evolutionary biology of this trait in a prominent model animal.

Method: …We derived all artificial selection lineages from a population of ~600 wild Drosophila melanogaster females caught in various locations in and around Hamilton, Ontario in late spring and early summer 2018…We mixed 3 F1 males and 3 F1 females from each of these isofemale lines together in 3 large populations. We then amplified these populations over one or 2 generations, generating a large total population size of ~6,000 flies, mixed among the three populations, and then randomly assigned flies to 12 separate lineages: 4 lineages to be selected for low sociability, 4 lineages to be selected for high sociability, and 4 control lab adaptation and domestication lineages

…We developed a novel ‘arena’ capable of both quantifying the sociability of groups of flies and allowing for the selection of flies based on their sociability (Figure 1A).

Figure 1: The arena used for quantification of and artificial selection on sociability. (A) The arena without the lid, showing the 8 compartments and an example arrangement of 16 flies. (B) The arena with the lid (note that the lid and swinging door were fully transparent, and opacity in the diagram is only for clarity). A foam plug at the central hole (not shown in the figure) allowed fly movement among the 8 compartments when at the top position, and locked flies within their compartment when in the bottom position.

Overview Of Artificial Selection Method: Overall, each generation, we tested 12 groups of 16 males and 12 groups of 16 females from each of the 8 selection lineages (4 low sociability, 4 high sociability). We selected 4 flies from each group of 16 flies to produce the next generation. In tests involving the low-sociability lineages, we chose the least sociable flies. In tests involving the high-sociability lineages, we chose the most sociable flies (see detailed methods below). We ran 2 experimental sessions per day over 2 days, with each session including 3 male groups and 3 female groups from each of the 8 lineages…At 1230h, we blocked the central area of each arena by pushing down the foam plug, sealing the flies into the compartment that they had settled in. At this point, we recorded the number of flies in each compartment of each arena. We then selected flies to produce the next generation for each lineage based on the number of flies in each compartment. We removed flies by rotating the lid so that the off-center hole was above a particular compartment, then rotating the plastic door so that the hole was uncovered, and aspirating the flies out. For the low-sociability lineages, we selected 4 flies per arena from compartments with the lowest numbers of flies, unless those numbers were 3 or more, in which case we took flies from other replicate arenas of that session with smaller groups. Similarly, for the high-sociability lineages, we selected 4 flies per arena from the compartment(s) with the highest number of flies, unless that number was 3 or less, in which case we took flies from larger groups in replicate arenas. The unselected flies from each arena were discarded. After each of the 4 selection sessions, we ended with 12 males and 12 females selected per lineage. We then placed the selected flies in sex-specific holding vials.

Figure 2: Divergence in selection treatments in sociability score over 25 generations. Mean ± SEM sociability scores across all selection lineages for low-sociability and high-sociability treatments in (A) females and (B) males. The same data are displayed by replicate lineages (error bars excluded for clarity) in (C) females and (D) males. Values statistically-significantly above 1 (dashed lines) indicate statistically-significant sociability.
Figure 4: Aggression frequency in (A) females and (B) males from the selection treatments after 25 generations of selection. Inner box plots show median, interquartile range (IQR), and whiskers up to 1.5 × IQR. Outer violin plots show the shape of the distribution of the data.