Phenotypic coupling of sleep and starvation resistance evolves in D. melanogaster

Abstract

Background: One hypothesis for the function of sleep is that it serves as a mechanism to conserve energy. Recent studies have suggested that increased sleep can be an adaptive mechanism to improve survival under food deprivation in Drosophila melanogaster. To test the generality of this hypothesis, we compared sleep and its plastic response to starvation in a temperate and tropical population of Drosophila melanogaster.

Results: We found that flies from the temperate population were more starvation resistant, and hypothesized that they would engage in behaviors that are considered to conserve energy, including increased sleep and reduced movement. Surprisingly, temperate flies slept less and moved more when they were awake compared to tropical flies, both under fed and starved conditions, therefore sleep did not correlate with population-level differences in starvation resistance. In contrast, total sleep and percent change in sleep when starved were strongly positively correlated with starvation resistance within the tropical population, but not within the temperate population. Thus, we observe unexpectedly complex relationships between starvation and sleep that vary both within and across populations. These observations falsify the simple hypothesis of a straightforward relationship between sleep and energy conservation. We also tested the hypothesis that starvation is correlated with metabolic phenotypes by investigating stored lipid and carbohydrate levels, and found that stored metabolites partially contributed towards variation starvation resistance.

Conclusions: Our findings demonstrate that the function of sleep under starvation can rapidly evolve on short timescales and raise new questions about the physiological correlates of sleep and the extent to which variation in sleep is shaped by natural selection.

Keywords: Behavior; Local adaptation; Plasticity; Sleep; Starvation resistance; Trade-off.

Fig. 1

ME flies are more starvation resistant compared to PC. a, ) Hour where half of the flies died of starvation for ME (white) and PC (gray) in (a) females and (c) males reared and maintained at 25 °C. Plot contains all data points from the 10 lines per population. b, d Boxplot of starvation resistance in hours for individual lines for ME and PC (b) females and (d) males. Solid line across the plot denotes the average value for each population

Fig. 2

ME flies sleep less at night and move more when awake compared to PC. a-b Mean sleep and standard error by Zeitgeber hour of (a) females and (b) males from ME (white) and PC (gray) populations at 25 °C during 12 h of light and 12 h dark periods, where night time is indicated by the gray shading of the plot. c-d Percentage of total sleep over a 24 h day (left), light period (middle) and night time (right) in ME and PC (c) females and (d) males. PC results are shaded in gray, and asterisks denote statistical significance after Bonferroni correction. Data points for daytime are denoted in circles and nighttime are in squares. e-f Average locomotor activity per minute when awake of (e) females and (f) males of ME and PC populations at 21 °C and 25 °C. Asterisks denote statistical significance after Bonferroni correction

Fig. 3

PC flies sleep more and engage in fewer locomotor activity when awake than ME under starvation. a-b Comparison of sleep patterns in ME flies when fed (black) and starved (gray). Data points for daytime are denoted in circles and nighttime are in squares. c-d Comparison of hourly sleep patterns of ME (white) and PC (gray) during starvation. Gray box denotes dark period, and Zeitgeber Hour denoted in these graphs start with when flies were placed on starvation media. e-f Total percent sleep when starved for ME and PC females and males. g-h Percent change in sleep upon starvation at the individual level for ME and PC flies. i-j Mean movement per minute when awake under starvation for (i) females and (j) males. Asterisks denote statistical significance after Bonferroni correction

Fig. 4

Coupling of starvation resistance and sleep is different in ME and PC populations. a-b Regression analysis of starvation resistance and percent of the 24 h day flies slept when starved for (a) ME and (b) PC. c-d Regression analysis of starvation resistance and percent change in sleep when starved for (c) ME and (d) PC flies. Each data point denotes the average of one of the 10 lines. Female data points are red circles with a solid regression line, while males are blue diamonds with a dashed line. The grey area denotes the confidence interval. * denotes p < 0.05 and ** denotes p < 0.01

Fig. 5

Stored metabolites are different in ME and PC populations. a Whole body glucose (mg glucose / mg protein) and (b) TGA (mg TGA / mg protein) in ME (while) and PC (gray) female and males. c-d Regression plots of starvation resistance and (c) glucose or (d) TGA levels in females. Black circles and solid lines denote the ME population and grey diamonds and dashed lines denote the PC population. The grey shaded area represents the confidence intervals of the regression. Panels a-b present distributions from each population / sex combination, while points in panel c and d represent the line means for each population. * denotes p < 0.05 and ** denotes p < 0.01