Working Title

Crowding effects in Drosophila melanogaster selected for early and late life fertility

Authors

Anastasia J. Shavrova & Adam K. Chippindale

Description

Rapid development can have benefits for many animals, especially in rapidly changin environments. However, trade-offs in a lack of plasticity can lead to an overall decline in fitness once the environment stabilizes. Additionally, populations are diverse and competitors are common. Using populations of flies selected for early and late life fertility, sharing a common ancestor, we compare how well they acclimate and survive in different crowding conditions. We then test the same populations in crowded conditions and a same-species competitor.

Hypotheses

Question1: Do early life fertility populations perform better than late life fertilty populations at higher densities?

Development time: Do populations change in development time to try and acclimate to higher densities?
- H1: early and late life fertility populations decrease development time to acclimate to a higher density
- H2: early and late life fertility populations increase development time to acclimate to a higher density
- H0: early and late life fertility populations do not change in development time
Viability: Do populations have a viability advantage due to their selected life history?
- H1: early life fertility populations have a higher viability than late life fertility populations
- H2: early life fertility populations have a lower viability than late life fertility populations
- H3: Both populations have equal viability.

Question 2: Do populations have an advantage over competitors due to their life history at high densities?

H1: Early and Late life fertility populations have an advantage over competitors at high densities
H2: Early life fertility populations have an advantage over competitors, while late life fertility populations do not
H3: Late life fertility populations have an advantage over competitors, while early life fertility populations do not
H0: neither population has an advantage over competitors at high densities

Brief Methods

Flies of two selection regimes: early life fertility (FAST) and late life fertility (SLOW), were cultured at densities 50, 100, 200, and 400 from egg to adult. Secondly, these selection regimes were then cultured in densities 100 and 400, but 50% of the density was occupied by a brown-eyed fly competitor. The adults were sexed and counted at emergence for a week. The time of egg hatch, pupation, and emergence was recorded.

Data Description

slctn = selection regime (FAST or SLOW)
pop = population replicate of selection regime (1,3,5)
rep = vial replicate (1,2,3,4,5)
dens = density of individuals (50, 100, 200, 400)
sex = sex of individual (female or male)
time = development time in days
indv = number of individuals
wght = weight of 5 males or females
eye = eye colour of fly (brown or red)
total = total number of individuals survived
brown = brown eye colour
red = red eye colour

Monoculture

Development

Summary

slctn	dens	sex	MeanDevelopment	MedianDevelopment	StandardDeviation	SampleSize	lower.ci	upper.ci
FAST	50	female	7.367652	7.346131	0.06932018	6	7.306890	7.428414
FAST	50	male	7.469649	7.470154	0.06617837	6	7.411641	7.527656
FAST	100	female	7.387900	7.398810	0.03711767	7	7.358200	7.417600
FAST	100	male	7.504233	7.514640	0.05756177	7	7.458174	7.550292
FAST	200	female	7.483724	7.492886	0.07189857	7	7.426193	7.541255
FAST	200	male	7.611373	7.633130	0.07710007	7	7.549680	7.673066
FAST	400	female	8.033767	8.027692	0.19023015	7	7.881551	8.185983
FAST	400	male	8.031056	8.026449	0.19748298	7	7.873037	8.189076
SLOW	50	female	9.025195	8.968697	0.10475238	6	8.933376	9.117015
SLOW	50	male	9.458077	9.346354	0.49417271	6	9.024915	9.891238
SLOW	100	female	9.522036	9.418239	0.25241078	7	9.320065	9.724007
SLOW	100	male	9.711858	9.782946	0.28554275	7	9.483376	9.940340
SLOW	200	female	10.620579	10.535714	0.56059342	7	10.172011	11.069147
SLOW	200	male	10.551032	10.374282	0.57944515	7	10.087379	11.014684
SLOW	400	female	11.967830	11.992477	0.20320653	7	11.805231	12.130429
SLOW	400	male	11.856797	11.707329	0.25130572	7	11.655711	12.057884

Graph

Modelling

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: time ~ dens * slctn * sex + (1 | pop) + (1 | rep)
##    Data: devo
## 
## REML criterion at convergence: 96.1
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -1.9450 -0.5520 -0.0449  0.3159  4.6129 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  rep      (Intercept) 0.000000 0.00000 
##  pop      (Intercept) 0.002989 0.05467 
##  Residual             0.078152 0.27956 
## Number of obs: 108, groups:  rep, 5; pop, 3
## 
## Fixed effects:
##                          Estimate Std. Error         df t value Pr(>|t|)    
## (Intercept)             7.1980657  0.0994760 44.9698193  72.360  < 2e-16 ***
## dens                    0.0019607  0.0004023 98.2177922   4.873 4.22e-06 ***
## slctnSLOW               1.5251877  0.1334451 98.2205678  11.429  < 2e-16 ***
## sexmale                 0.1489405  0.1333399 98.0007755   1.117    0.267    
## dens:slctnSLOW          0.0063847  0.0005692 98.3221146  11.216  < 2e-16 ***
## dens:sexmale           -0.0003309  0.0005686 98.0007755  -0.582    0.562    
## slctnSLOW:sexmale       0.2118209  0.1885711 98.0007755   1.123    0.264    
## dens:slctnSLOW:sexmale -0.0010304  0.0008041 98.0007755  -1.281    0.203    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) dens   slSLOW sexmal dn:SLOW dns:sx sSLOW:
## dens        -0.779                                           
## slctnSLOW   -0.671  0.581                                    
## sexmale     -0.670  0.580  0.500                             
## dns:slcSLOW  0.551 -0.707 -0.822 -0.410                      
## dens:sexmal  0.550 -0.707 -0.410 -0.821  0.499               
## slctnSLOW:s  0.474 -0.410 -0.707 -0.707  0.580   0.581       
## dns:slSLOW: -0.389  0.500  0.580  0.581 -0.706  -0.707 -0.821
## convergence code: 0
## boundary (singular) fit: see ?isSingular

Summary

The SLOW selection lines significantly increase as density increases, while the FAST selection lines have a slight increase as density increases but this change is not significant. There are no difference between sex in development time for FAST and SLOw selection regimes.

Viability

Summary

slctn	dens	sex	MeanViability	MedianViability	StandardDeviation	SampleSize	lower.ci	upper.ci
FAST	50	female	20.66667	20	5.108350	15	17.99075	23.34258
FAST	50	male	18.26667	17	5.311264	15	15.48446	21.04888
FAST	100	female	43.33333	43	6.019809	15	40.17996	46.48670
FAST	100	male	42.60000	40	7.585701	15	38.62637	46.57363
FAST	200	female	78.80000	77	10.577604	15	73.25911	84.34089
FAST	200	male	76.40000	80	13.870833	15	69.13401	83.66599
FAST	400	female	144.33333	151	20.194294	15	133.75492	154.91175
FAST	400	male	136.40000	144	30.279177	15	120.53880	152.26120
SLOW	50	female	21.40000	21	5.369224	15	18.58743	24.21257
SLOW	50	male	22.86667	23	5.180550	15	20.15293	25.58040
SLOW	100	female	53.66667	52	8.682220	15	49.11864	58.21469
SLOW	100	male	49.40000	44	13.751883	15	42.19632	56.60368
SLOW	200	female	74.00000	76	8.684962	15	69.45054	78.54946
SLOW	200	male	70.33333	69	11.763543	15	64.17121	76.49545
SLOW	400	female	73.40000	67	22.468390	15	61.63034	85.16966
SLOW	400	male	82.46667	83	18.403675	15	72.82623	92.10710

Graph

Modelling

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: indv ~ dens * slctn * sex + (1 | pop) + (1 | rep)
##    Data: viab
## 
## REML criterion at convergence: 2022.2
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.1910 -0.5233  0.0182  0.6404  2.6935 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  rep      (Intercept)   0.2891  0.5377 
##  pop      (Intercept)  12.7293  3.5678 
##  Residual             258.2072 16.0688 
## Number of obs: 240, groups:  rep, 5; pop, 3
## 
## Fixed effects:
##                         Estimate Std. Error        df t value Pr(>|t|)    
## (Intercept)              6.50725    4.12604  19.37115   1.577    0.131    
## dens                     0.34814    0.01548 226.00011  22.496  < 2e-16 ***
## slctnSLOW               25.93333    5.04445 226.00011   5.141 5.91e-07 ***
## sexmale                  0.08116    5.04445 226.00011   0.016    0.987    
## dens:slctnSLOW          -0.22453    0.02189 226.00011 -10.259  < 2e-16 ***
## dens:sexmale            -0.01839    0.02189 226.00011  -0.840    0.402    
## slctnSLOW:sexmale       -4.78551    7.13393 226.00011  -0.671    0.503    
## dens:slctnSLOW:sexmale   0.04694    0.03095 226.00011   1.517    0.131    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) dens   slSLOW sexmal dn:SLOW dns:sx sSLOW:
## dens        -0.703                                           
## slctnSLOW   -0.611  0.575                                    
## sexmale     -0.611  0.575  0.500                             
## dns:slcSLOW  0.497 -0.707 -0.813 -0.407                      
## dens:sexmal  0.497 -0.707 -0.407 -0.813  0.500               
## slctnSLOW:s  0.432 -0.407 -0.707 -0.707  0.575   0.575       
## dns:slSLOW: -0.352  0.500  0.575  0.575 -0.707  -0.707 -0.813

Summary

There are no significant viability differences between selection regimes for densities 100-200. For the highest density of 400, SLOW viability is significantly poorer than the FAST viability. There are no sex differences in viability for each selection regime.

Weight

Summary

slctn	dens	sex	MeanWeight	MedianWeight	StandardDeviation	SampleSize	lower.ci	upper.ci
FAST	50	female	1.0918125	1.09100	0.05923425	16	1.0618358	1.1217892
FAST	50	male	0.8347619	0.86900	0.15418525	21	0.7671872	0.9023366
FAST	100	female	1.0348261	1.06850	0.08792833	46	1.0091352	1.0605169
FAST	100	male	0.8227872	0.81700	0.06570824	47	0.8037985	0.8417760
FAST	200	female	0.8807059	0.89350	0.09174642	68	0.8587370	0.9026748
FAST	200	male	0.7193158	0.72650	0.08104323	76	0.7009740	0.7376576
FAST	400	female	0.6449717	0.65000	0.07335737	106	0.6309402	0.6590032
FAST	400	male	0.5385096	0.53300	0.07015157	104	0.5249616	0.5520576
SLOW	50	female	1.8871842	1.91300	0.11926642	19	1.8320859	1.9422825
SLOW	50	male	1.4151786	1.42475	0.05697287	14	1.3842078	1.4461494
SLOW	100	female	1.6990417	1.73400	0.12612202	48	1.6629840	1.7350994
SLOW	100	male	1.2640306	1.30250	0.08767812	49	1.2392263	1.2888349
SLOW	200	female	1.5691190	1.54150	0.11563443	63	1.5403353	1.5979028
SLOW	200	male	1.1923934	1.18650	0.09073827	61	1.1694335	1.2153534
SLOW	400	female	1.3605000	1.35725	0.11499233	18	1.3058361	1.4151639
SLOW	400	male	1.0379250	1.04075	0.08880794	20	0.9979921	1.0778579

Graph

Modelling

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: wght ~ dens * slctn * sex + (1 | pop) + (1 | rep)
##    Data: weight
## 
## REML criterion at convergence: -1398.5
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -5.0287 -0.5310  0.0625  0.5617  3.5478 
## 
## Random effects:
##  Groups   Name        Variance  Std.Dev.
##  rep      (Intercept) 4.072e-04 0.020180
##  pop      (Intercept) 3.388e-05 0.005821
##  Residual             8.432e-03 0.091827
## Number of obs: 776, groups:  rep, 5; pop, 3
## 
## Fixed effects:
##                          Estimate Std. Error         df t value Pr(>|t|)    
## (Intercept)             1.147e+00  1.630e-02  2.994e+01  70.365  < 2e-16 ***
## dens                   -1.265e-03  4.508e-05  7.659e+02 -28.050  < 2e-16 ***
## slctnSLOW               7.136e-01  1.991e-02  7.646e+02  35.849  < 2e-16 ***
## sexmale                -2.478e-01  1.810e-02  7.632e+02 -13.694  < 2e-16 ***
## dens:slctnSLOW         -8.635e-05  8.710e-05  7.633e+02  -0.991    0.322    
## dens:sexmale            3.654e-04  6.260e-05  7.630e+02   5.836 7.89e-09 ***
## slctnSLOW:sexmale      -2.342e-01  2.803e-02  7.635e+02  -8.355 3.08e-16 ***
## dens:slctnSLOW:sexmale  9.773e-05  1.224e-04  7.640e+02   0.799    0.425    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) dens   slSLOW sexmal dn:SLOW dns:sx sSLOW:
## dens        -0.717                                           
## slctnSLOW   -0.529  0.578                                    
## sexmale     -0.583  0.641  0.477                             
## dns:slcSLOW  0.368 -0.510 -0.856 -0.330                      
## dens:sexmal  0.513 -0.714 -0.419 -0.887  0.366               
## slctnSLOW:s  0.376 -0.414 -0.708 -0.646  0.605   0.573       
## dns:slSLOW: -0.264  0.367  0.608  0.454 -0.709  -0.512 -0.860

Summary

Females lose weight significantly faster than males as density increases, yet at the same rate between selection regimes (i.e SLOW females lose weight at the same rate as FAST females).

Competition

Development

Summary

slctn	dens	eye	sex	MeanDevelopment	MedianDevelopment	StandardDeviation	SampleSize	lower.ci	upper.ci
FAST	100	brown	female	8.986923	8.880000	0.4671614	13	8.722602	9.251245
FAST	100	brown	male	9.153846	9.000000	0.4995755	13	8.871185	9.436508
FAST	100	red	female	7.740000	7.600000	0.3531053	13	7.540212	7.939788
FAST	100	red	male	7.837692	7.650000	0.3911128	13	7.616399	8.058985
FAST	400	brown	female	10.545120	10.573574	0.4006784	14	10.327309	10.762932
FAST	400	brown	male	10.428700	10.401663	0.4693493	14	10.173559	10.683841
FAST	400	red	female	8.097449	8.120343	0.3374882	14	7.913988	8.280909
FAST	400	red	male	8.160666	8.118217	0.3152499	14	7.989294	8.332038
SLOW	100	brown	female	9.307857	9.285000	0.4913902	14	9.040734	9.574980
SLOW	100	brown	male	9.543571	9.470000	0.5200090	14	9.260891	9.826252
SLOW	100	red	female	9.240000	9.165000	0.3577709	14	9.045514	9.434486
SLOW	100	red	male	9.400714	9.365000	0.3493833	14	9.210787	9.590641
SLOW	400	brown	female	10.592896	10.611239	0.3487771	14	10.403299	10.782493
SLOW	400	brown	male	10.642534	10.658523	0.4143518	14	10.417290	10.867778
SLOW	400	red	female	11.788047	11.708848	0.3708369	14	11.586458	11.989636
SLOW	400	red	male	11.620023	11.627692	0.3571899	14	11.425853	11.814194

Graph

Modelling

## Linear mixed model fit by REML. t-tests use Satterthwaite's method [
## lmerModLmerTest]
## Formula: time ~ dens * eye * slctn * sex + (1 | pop) + (1 | rep)
##    Data: comp
## 
## REML criterion at convergence: 271
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -2.1526 -0.6885 -0.0446  0.6727  3.5487 
## 
## Random effects:
##  Groups   Name        Variance Std.Dev.
##  rep      (Intercept) 0.02917  0.1708  
##  pop      (Intercept) 0.09639  0.3105  
##  Residual             0.10534  0.3246  
## Number of obs: 220, groups:  rep, 5; pop, 3
## 
## Fixed effects:
##                                 Estimate Std. Error         df t value Pr(>|t|)
## (Intercept)                    8.425e+00  2.308e-01  5.085e+00  36.499 2.37e-07
## dens                           5.253e-03  4.169e-04  1.979e+02  12.601  < 2e-16
## eyered                        -8.467e-01  1.746e-01  1.979e+02  -4.849 2.50e-06
## slctnSLOW                      6.356e-01  1.736e-01  1.981e+02   3.662 0.000321
## sexmale                        2.614e-01  1.746e-01  1.979e+02   1.497 0.135983
## dens:eyered                   -4.002e-03  5.893e-04  1.979e+02  -6.792 1.26e-10
## dens:slctnSLOW                -9.697e-04  5.840e-04  1.979e+02  -1.661 0.098382
## eyered:slctnSLOW               3.578e-01  2.427e-01  1.979e+02   1.474 0.142008
## dens:sexmale                  -9.445e-04  5.893e-04  1.979e+02  -1.603 0.110596
## eyered:sexmale                -1.522e-01  2.469e-01  1.979e+02  -0.616 0.538369
## slctnSLOW:sexmale              3.637e-02  2.427e-01  1.979e+02   0.150 0.881040
## dens:eyered:slctnSLOW          8.212e-03  8.256e-04  1.979e+02   9.947  < 2e-16
## dens:eyered:sexmale            8.296e-04  8.334e-04  1.979e+02   0.995 0.320761
## dens:slctnSLOW:sexmale         3.242e-04  8.256e-04  1.979e+02   0.393 0.694972
## eyered:slctnSLOW:sexmale       1.247e-01  3.432e-01  1.979e+02   0.363 0.716683
## dens:eyered:slctnSLOW:sexmale -1.305e-03  1.168e-03  1.979e+02  -1.118 0.265042
##                                  
## (Intercept)                   ***
## dens                          ***
## eyered                        ***
## slctnSLOW                     ***
## sexmale                          
## dens:eyered                   ***
## dens:slctnSLOW                .  
## eyered:slctnSLOW                 
## dens:sexmale                     
## eyered:sexmale                   
## slctnSLOW:sexmale                
## dens:eyered:slctnSLOW         ***
## dens:eyered:sexmale              
## dens:slctnSLOW:sexmale           
## eyered:slctnSLOW:sexmale         
## dens:eyered:slctnSLOW:sexmale    
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Summary

Competitor and the SLOW selection regime increase in development time as density increases, where SLOW selection regime increases more so than the competitor. FAST selection regime does not change in development time.

Viability

Summary

slctn	dens	eye	sex	MeanViability	MedianViability	StandardDeviation	SampleSize	lower.ci	upper.ci
FAST	100	brown	female	21.92308	22.0	4.009604	13	19.65443	24.19172
FAST	100	brown	male	23.38462	23.0	3.404371	13	21.45841	25.31082
FAST	100	red	female	22.00000	21.0	3.316625	13	20.12344	23.87656
FAST	100	red	male	18.76923	20.0	2.554533	13	17.32387	20.21459
FAST	400	brown	female	70.71429	70.5	7.415458	14	66.68320	74.74537
FAST	400	brown	male	67.78571	67.5	7.051225	14	63.95262	71.61880
FAST	400	red	female	54.00000	55.0	8.548504	14	49.35298	58.64702
FAST	400	red	male	49.57143	48.0	8.308608	14	45.05482	54.08804
SLOW	100	brown	female	25.28571	24.5	3.729891	14	23.25812	27.31331
SLOW	100	brown	male	19.35714	19.0	3.499608	14	17.45473	21.25955
SLOW	100	red	female	20.14286	20.0	4.258153	14	17.82810	22.45762
SLOW	100	red	male	22.57143	23.0	4.089655	14	20.34827	24.79459
SLOW	400	brown	female	53.92857	51.5	9.714892	14	48.64750	59.20965
SLOW	400	brown	male	49.35714	48.5	5.930458	14	46.13331	52.58098
SLOW	400	red	female	72.85714	74.0	11.461354	14	66.62668	79.08761
SLOW	400	red	male	61.64286	61.5	10.717399	14	55.81681	67.46890

Graph

Modelling

## Generalized linear mixed model fit by maximum likelihood (Laplace
##   Approximation) [glmerMod]
##  Family: binomial  ( logit )
## Formula: cbind(red, brown) ~ dens * sex * slctn + total + (1 | rep) +  
##     (1 | pop)
##    Data: comp2
## 
##      AIC      BIC   logLik deviance df.resid 
##    650.1    680.7   -314.0    628.1      109 
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -3.0885 -0.4147  0.0000  0.5567  2.6578 
## 
## Random effects:
##  Groups Name        Variance Std.Dev. 
##  rep    (Intercept) 0.00e+00 0.000e+00
##  pop    (Intercept) 1.55e-18 1.245e-09
## Number of obs: 120, groups:  rep, 5; pop, 3
## 
## Fixed effects:
##                       Estimate Std. Error z value Pr(>|z|)    
## (Intercept)          0.0035419  0.1210526   0.029 0.976658    
## dens                -0.0023280  0.0007573  -3.074 0.002110 ** 
## sexM                -0.2846303  0.1608583  -1.769 0.076819 .  
## slctnSLOW           -0.5070050  0.1551649  -3.268 0.001085 ** 
## total                0.0027041  0.0013154   2.056 0.039806 *  
## dens:sexM            0.0006079  0.0004596   1.323 0.185997    
## dens:slctnSLOW       0.0027204  0.0004455   6.106 1.02e-09 ***
## sexM:slctnSLOW       0.8185333  0.2229594   3.671 0.000241 ***
## dens:sexM:slctnSLOW -0.0021405  0.0006409  -3.340 0.000838 ***
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) dens   sexM   slSLOW total  dns:sM d:SLOW sM:SLO
## dens        -0.024                                                 
## sexM        -0.646  0.270                                          
## slctnSLOW   -0.660  0.302  0.501                                   
## total       -0.375 -0.906  0.004 -0.021                            
## dens:sexM    0.597 -0.294 -0.927 -0.464  0.000                     
## dns:slcSLOW  0.599 -0.346 -0.461 -0.925  0.047  0.497              
## sxM:slcSLOW  0.464 -0.193 -0.719 -0.693 -0.003  0.667  0.641       
## dns:sM:SLOW -0.426  0.210  0.663  0.640 -0.001 -0.715 -0.691 -0.925
## convergence code: 0
## boundary (singular) fit: see ?isSingular

Summary

FAST selection regime is outcompeted by the brown eyed competitor, this is especially so for the FAST males. However, the SLOW selection regime is a much better competitor at higher densities for both sexes.

Crowding effects of early and late life fertility flies

Anastasia Shavrova

2020-08-20

Working Title

Authors

Description

Hypotheses

Brief Methods

Data Description

Monoculture

Development

Summary

Graph

Modelling

Viability

Summary

Graph

Modelling

Weight

Summary

Graph

Modelling

Competition

Development

Summary

Graph

Modelling

Viability

Summary

Graph

Modelling