Neuroscience

Selection for male aggression is associated with changes in reproductive traits, chemical signaling and lifespan in Drosophila melanogaster

Anthony Defert
Romane Gout
Gaelle Pennot
Fanny Jamme
Amandine Castex
Anissa Handjar
Thomas Guilleman
Jean-Christophe Billeter
Séverine Trannoy author has email address

Research Center on Animal Cognition (CRCA), Center for Integrative Biology, Toulouse University, CNRS, UPS, Toulouse, France
Institute for Zoology, Halle-Wittenberg University, Halle, Germany
Groningen Institute for Evolutionary Life Sciences, University of Groningen, Groningen, Netherlands

https://doi.org/10.7554/eLife.107559.2

Open access
Copyright information

Figures and data

Courtship and aggressive behaviors in Cs, Bully A, Bully B and Bully C lines.
(A) Schema of the experimental set up to record aggressive and courtship behaviors. Couples of flies and presence of food cup are mentioned according to the figure of this article. (B) Latency to lunge of Cs, Bully A, Bully B and Bully C males in aggression assay with a territory against a male of the same line (n_Cs=29, n_BullyA=24, n_BullyB=12, n_BullyC=11, LM and permutation test (Monte-Carlo test): P=0.136). (C) Number of lunges of Cs, Bully A, Bully B and Bully C male pairs in aggression assays (n_Cs=29, n_BullyA=24, n_BullyB=12, n_BullyC=11, GLM quasi-Poisson and type II analysis of deviance table (Likelihood Ratio Test): P=1.622×10⁻¹⁰, LRχ²=48.556, post-hoc: P_{Cs_BullyA}<0.001, z ratio= −4.495, P_{Cs_BullyB}<0.001, z.ratio= −5.911, P_{Cs_BullyC}=0.018, z.ratio= −2.836, P_{BullyA_BullyB}=0.042, z ratio= −2.308, P_{BullyA_BullyC}=0.232, z ratio= 1.195, P_{BullyB_BullyC}=0.018, z ratio=2.797). (D) Number of boxing events of Cs, Bully A, Bully B and Bully C male pairs in aggression assays (n_Cs=29, n_BullyA=24, n_BullyB=12, n_BullyC=11, GLM Poisson and type II analysis of deviance table (Likelihood Ratio Test): P=2.2×10⁻¹⁶, LRχ²=2511.9, post-hoc: P_{Cs_BullyA}<0.001, z ratio= −6.819, P_{Cs_BullyB}<0.001, z.ratio= −6.839, P_{Cs_BullyC}<0.001, z.ratio= −3.649, P_{BullyA_BullyB}=0.728, z ratio= −0.347, P_{BullyA_BullyC}<0.001, z ratio=5.392, P_{BullyB_BullyC}<0.001, z ratio=5.366). (E) Non-competitive mating success of Cs, Bully A, Bully B and Bully C males paired with a virgin Cs female (n_Cs=37, n_BullyA=31, n_BullyB=33, n_BullyC=21, GLM binomial and type II analysis of deviance table (Likelihood Ratio Test): P=0.012, LRχ²=11.040, post-hoc: P_{Cs_BullyA}=0.934, z ratio=0.728, P_{Cs_BullyB}=0.099, z.ratio=2.395, P_{Cs_BullyC}=0.868, z.ratio=1.060, P_{BullyA_BullyB}=0.206, z ratio=2.043, P_{BullyA_BullyC}=0.934, z ratio=0.406, P_{BullyB_BullyC}=0.522, z ratio=-1.513). (F) Proportion of time spent by Cs and Cs, Bully A, Bully B and Bully C males doing Unilateral Wing Extension (UWE) toward a Cs virgin female in non-competitive courtship assays (n_Cs=37, n_BullyA=31, n_BullyB=33, n_BullyC=21, LM and type II Anova table: P=0.088, F=2.235). (G) Latency to court (time difference between first interaction and first UWE) of Cs, Bully A, Bully B and Bully C males paired with a virgin Cs female (n_Cs=37, n_BullyA=31, n_BullyB=33, n_BullyC=21, LM and type II Anova table: P=0.014, F=3.707, post-hoc: P_{Cs_BullyA}=0.338, t.ratio= −1.740, P_{Cs_BullyB}=0.433, t.ratio= 1.469, P_{Cs_BullyC}=0.524, t.ratio= −1.127, P_{BullyA_BullyB}=0.015, t.ratio=3.101, P_{BullyA_BullyC}=0.683, t.ratio=0.661, P_{BullyB_BullyC}=0.099, t.ratio= −2.364). (H) Latency to mate (time difference between first interaction and start of the mating) of Cs, Bully A, Bully B and Bully C males paired with a virgin Cs female (n_Cs=36, n_BullyA=29, n_BullyB=24, n_BullyC=19, LM and type II Anova table: P=8.527×10⁻⁴, F=5.970, post-hoc: P_{Cs_BullyA}=0.482, t.ratio= −1.413, P_{Cs_BullyB}<0.001, t.ratio= −4.191, P_{Cs_BullyC}=0.581, t.ratio= −1.062, P_{BullyA_BullyB}=0.038, t.ratio= −2.725, P_{BullyA_BullyC}=0.863, t.ratio=0.173, P_{BullyB_BullyC}=0.041, t.ratio= 2.615). (I) Mating duration of Cs, Bully A, Bully B and Bully C males paired with a virgin Cs female (n_Cs=36, n_BullyA=29, n_BullyB=24, n_BullyC=19, LM and type II Anova table: P=8.527×10⁻⁴, F=5.970, post-hoc: P_{Cs_BullyA}=0.482, t.ratio= −1.413, P_{Cs_BullyB}<0.001, t.ratio= −4.191, P_{Cs_BullyC}=0.581, t.ratio= −1.062, P_{BullyA_BullyB}=0.038, t.ratio= −2.725, P_{BullyA_BullyC}=0.863, t.ratio=0.173, P_{BullyB_BullyC}=0.041, t.ratio= 2.615). For all graphs, different letters mean a statistical difference. Statistical details are given in Figure 1-table supplement 1.

Bully males exhibit reduced mating success and allocate less time to mating.
(A) In competitive courtship assays, the mating success of Cs and Bully males was scored when paired with a single virgin Cs female (n_Cs=19, n_Bully=19, GLM binomial and type II analysis of deviance table (Likelihood Ratio): Test, P=0.329, LRχ²=0.951). (B) In non-competitive courtship assays, the mating success of Cs and Bully males paired with a virgin Cs female were analyzed (n_Cs=64, n_Bully=41, GLM binomial and type II analysis of deviance table (Likelihood Ratio Test): P=0.037, LRχ²=4.347). (C) Proportion of time spent by Cs and Bully males doing Unilateral Wing Extension (UWE) toward a Cs virgin female in non-competitive courtship assays (n_Cs=64, n_Bully=41, Fisher Pitman permutation test: P=0.877, t= −0.131). (D) Latency to court (time difference between first interaction and first UWE) of Cs and Bully males paired with a virgin Cs female (n_Cs=64, n_Bully=41, Fisher Pitman permutation test: P=0.319, Z=1.043). (E) Latency to mate (time difference between first interaction and start of the mating) of Cs and Bully males paired with a virgin Cs female (n_Cs=61, n_Bully=34, Fisher Pitman permutation test: P=0.266, Z=1.170). (F) Mating duration of Cs and Bully males paired with a virgin Cs female (n_Cs=61, n_Bully=34, T-test: P=6.009×10⁻¹⁰, t= −6.917) or (G) a virgin Cs decapitated female (n_Cs=26, n_Bully=20, T-test: P=0.022, t= −2.367). (H) Mating duration of Cs and Bully males that were previously raised in group of 10 males or in social isolation (n_Cs-isolated=22, n_Cs-group=32, n_{Bully-isolated}=23, n_Bully-group=25, LM and type II ANOVA: F-test, P=7.582×10⁻¹¹, F=21.669, post-hoc: P_{Cs-isolated_Cs-group}=0.013, F=6.583, P_{Bully-isolated_Bully-group}=2.532×10⁻⁴, F=15.731, P_{Cs-isolated_Bully-isolated}=4.263×10⁻⁶, F=27.697, P_{Cs-group_Bully-group}=1.884×10⁻⁴, F=16.032). For all graphs, stars indicate significant differences (* P<0.05, ** P< 0.01, *** P<0.001) and statistical details are given in Figure 1-table supplement 1.

Hyper-aggressive males exhibit changes in mate-guarding features.
(A, B, F, G). Mating duration of Cs and Bully males paired with a virgin Cs female (A: n_Cs=21, n_Bully=22, T-test: P=2.912×10⁻¹¹, t= −8.759, B - n_Cs=24, n_Bully=24, T-test: P=5.425×10⁻⁷, t= −5.818, F - n_Cs=25, n_Bully=25, T-test: P=7.139×10⁻⁴, t= −3.617, G - n_Cs=30, n_Bully=24, T-test: P=8.743×10⁻⁵, t= −4.255). (A’) After mating ended, the proportion of time spent by each male to court performing UWE towards the just mated female was scored for 10 minutes (n_Cs=21, n_Bully=22, Fisher Pitman permutation test: P<0.001, Z=3.239). (B’) The proportion of time spent by the males performing UWE an unfamiliar mated female was scored for 10 minutes (n_Cs=24, n_Bully=24, Fisher Pitman permutation test: P=0.007, Z= −2.679). (C) Total number of mating reached by Cs and Bully males over six consecutive mating with virgin Cs females (type II analysis of deviance table (n_Cs=18♂, n_Bully=18♂, Wald Chi-square test): P=0.659, χ²=0.195). (D) Mating duration of Cs and Bully males over six consecutive mating with virgin Cs females (n_Cs=18♂, n_Bully=18♂, LMM and Satterthwaite T-test: P_interaction=0.001, t= −3.261). A significant interaction between males’ genotype and number of mating was found, indicating that Cs decreased their mating duration with the number of mating, while Bully males did not. (E) Proportion of mating that gave rise to progeny in consecutive mating assay (n_Cs=18♂, n_Bully=18♂, GLMM and Wald Chi-square test: P_{male_genotype}=0.279, χ²=1.173, P_mating=0.847, χ²=2.019). (F’) Proportion of time spent by naïve Cs and Bully males to display UWE either toward Cs-mated or Bully-mated decapitated female over an observation period of 2 minutes (n_Cs-Cs-mated=12, n_{Cs-Bully-mated}=15, n_{Bully-Cs-mated}=9, n_{Bully-Bully-mated}=10, LMM and permutation test (Monte-Carlo test): P_{male_genotype}<0.001, P_{mated_female}=0.457). (G’) Proportion of Cs-mated and Bully-mated females that re-mated 12 or 14 days after a first mating (n_Cs-12days=19, n_Bully-12days=15, n_Cs-14days=11, n_Bully-14days=9, GLM binomial and type II analysis of deviance table (Likelihood Ratio Test): P_{mated_female}=3.639×10⁻⁴, LRχ²=12.708, P_day=0.385, LRχ²=0.756). For graphs A, A’, B, B’, F, F’, G and G’, stars indicate significant differences (* P < 0.05, ** P < 0.01, *** P < 0.001) and statistical details are given in Figure 1-table supplement 1.

Comparison of CHCs concentrations between naïve Cs and Bully males.
Mean ± SE concentrations are provided in ng for each compound and each group of compounds. In statistics columns, stars represent significant differences (* P < 0.05, ** P < 0.01, *** P < 0.001). Statistical details 986 are given in Figure 1-table supplement 1.

Hyper-aggressive males exhibit altered CHCs profiles and transfer reduced amounts of mate-guarding pheromones to females.
(A) PCA analysis on mean concentration for each Cuticular Hydrocarbon (CHCs) compound of naïve Cs and Bully. Their CHCs profiles were found similar (n_Cs=10♂, n_Bully=10♂, PerMANOVA: P=0.002, F=6.789, R²=0.274). (B) Relative concentrations of Bully’s CHCs were calculated with respect to those of Cs males. The CHCs levels in Cs males were normalized to 1, and Bully’s CHC concentrations were expressed relative to this reference (n_Cs=10♂, n_Bully=10♂, LMM and permutation test (Monte-Carlo test): P=0.009, post-hoc: P_Alkanes=0.036, P_Monoenes=0.002, P_{Methyl alkanes}=0.048, P_nC21<0.001, P_C22:1(9)=0.892, P_C22:1(7)=0.568, P_cVA=0.731, P_nC22=0.854, P_C23:1(9)=0.175, P_C23:1(7)=0.062, P_C23:1(5)<0.001, P_nC23=0.014, P_C24:1(9)=0.345, P_C24:1(7<0.001, P_24:1(5)=0.015, P_nC24=0.002, P_2MeC24=0.548, P_C25:1(9)=0.524, P_C25:1(7)<0.001, P_C25:1(5)<0.001, P_nC25<0.001, P_2MeC26=0.159, P_C27:1(7)<0.001, P_nC27=0.078, P_2MeC28=0.003, P_nC29=0.005). (C) PCA analysis on mean concentration for each CHCs compound of virgin, Cs-mated and Bully-mated females (n_virgin=10♀, n_Cs-mated=16♀, n_Bully-mated=14♀, PerMANOVA: P=0.017, F=2.594, R²=0.123, post-hoc: P_{virgin_Cs-mated}=0.001, F=19.434, R²=0.447, P_{virgin_Bully-mated}=0.034, F=2.266, R²=0.093, P_{Cs-mated_Bully-mated}=0.746, F=0.428, R²=0.015). (D) Relative concentrations of CHCs compounds in mated females compared to virgin females. CHCs levels in virgin females were normalized to 1, and the concentrations in mated females were expressed relative to this baseline (LMM and permutation test (n_virgin=10♀, n_Cs-mated=16♀, n_Bully-mated=14♀, Monte-Carlo test: P=0.618) Even though the p-value was not significant, post-hoc: P_Alkanes=0.705, P_Monoenes=0.294, P_Dienes=0.715, P_{Methyl alkanes}=0.750, P_nC21=0.158, Pn_C22=0.927, P_7,11C23:2<0.001, P_2MeC22<0.001, P_C23:1(9)=0.023, P_C23:1(7)=0.001, P_C23:1(5)=0.016, P_nC23=0.908, P_nC24=0.899, P_7,11C25:2<0.001, P_2MeC24<0.001, P_C25:1(9)=0.832, P_C25:1(7)=0.047, P_C25:1(5)<0.001, P_nC25=0.769, P_9,13C27:2<0.001, P_7,11C27:2=0.747, P_2MeC26=0.884, P_C27:1(9)=0.016, P_C27:1(7)=0.625, P_C27:1(5)<0.001, P_nC27=0.367, P_9,13C29:2<0.001, P_7,11C23:2=0.899, P_2MeC28=0.413, P_C29:1(9)<0.001, P_C29:1(7)<0.001, P_nC29=0.022). Significant differences are indicated by letters: a indicates a significant difference between virgin and Cs-mated females; b indicates a significant difference between virgin and Bully-mated females; and c indicates a significant difference between Cs- and Bully-mated females (a P < 0.05; aa P < 0.01; aaa P < 0.001). (E) Mean ± SE concentration in ng of CHCs compounds that were not present in virgin females (n_virgin=10♀, n_Cs-mated=16♀, n_Bully-mated=14♀, P_C23:1(6)<0.001, P_C25:1(6)<0.001, P_cVA=0.001). (F) Mating duration of Cs and Bully males with a Cs female (n_Cs-mated=17, n_Bully-mated=14, T-test: P=0.003, t= −3.359). (F’) cVA concentration measured within Cs- and Bully-mated genitalia tract right after mating (n_Cs-mated=17, n_Bully-mated=14, Fisher Pitman permutation test: P=0.001, Z= −2.390). For graphs E, F and F’, stars indicate significant differences (* P < 0.05, ** P < 0.01, *** P < 0.001) and statistical details are given in Figure 1-table supplement 1.

Comparison of CHCs concentrations among virgin, Cs-mated, and Bully-mated females.
Mean ± SE concentrations are provided in ng for each compound and each group of compounds. In the statistics columns, stars indicate significant differences (* P < 0.05, ** P < 0.01, *** P < 0.001). Statistical details are given in Figure 1-table supplement 1.

Aggressive males offset their reduced reproductive success with a longer lifespan and preserved reproductive success in later life.
(A) Survival curves of Cs (yellow) and homozygous (green) and heterozygous (red) Bully males (Weibull regression model (n =50, n =50, n =51, Likelihood Ratio Test): χ²=23.91, p=2.6×10⁻⁵, post-hoc: P_{Cs – Bully_homo}<0.001, t.ratio= 4.061, P_{Cs – Bully_hetero}< 0.001, t.ratio=5.088, P_{Bully_homo – Bully_hetero}=0.318, t.ratio = −1.001). Cs males had a median survival (EC₅₀) of 57 days, with the last individual dying at day 90 (yellow), while progeny from homozygous Bully males showed an EC₅₀ of 82 days, with the final death at day 101 (green) and heterozygous Bully males showed an EC₅₀ of 79 days, with the final death at day 106 (red). (B) Survival curves of group of 10 Cs and 10 homozygous Bully males raised in group with 12 Cs females. Bully males had a median survival (EC₅₀) of 56 days, with the last individual dying at day 71 (red line), while Cs males showed an EC₅₀ of 43 days, with the final death at day 67 (yellow line) (n_Cs=50, n_Bully-hetero=50, Cox proportional hazards regression model with stratification by group (Likelihood Ratio Test): P=7×10⁻⁵, χ²=15.94). Statistical details are given in Figure 1-table supplement. (C) Mating duration of Cs and Bully males with virgin Cs females across their lifespan (n_Cs-J0=24, n_Bully-J14=23, LMM and Satterthwaite T-test: P_interaction =3.360×10⁻⁵, t=4.227). A significant interaction between male genotype and age was observed: Bully males maintained a stable mating duration over time, whereas Cs males showed an increase in mating duration with successive mating events. (D) Number of mating throughout the lifespan of Cs and Bully males (n_Cs-J0=24, n_Bully-J14=23, GLM binomial and Wald Z-test: P_interaction=0.039, Z= −2.068). Numbers in bars indicate the proportion of mating. Arrows denote the last mating event and the number of offspring produced. Statistical details are given in Figure 1-table supplement 1. Table 1: Comparison of CHCs concentrations between naïve Cs and Bully males. Mean ± SE concentrations are provided in ng for each compound and each group of compounds. In statistics columns, stars represent significant differences (* P < 0.05, ** P < 0.01, *** P < 0.001). Statistical details are given in Figure 1-table supplement 1.

Sign up for email alerts