Body size is a trait that shapes many aspects of a species’ development and evolution. Larger body size is often beneficial in animals, but it can also be associated with life history costs in natural systems. Similarly, miniaturization, the evolution of extremely small adult body size, is found in every major animal group, yet carries its own life history trade-offs. Given that these effects can depend on an animal’s environment and life stage and have mainly been studied in species that are already specialized for their size, the life history changes associated with evolutionary shifts in body size warrant additional investigation. Here, we used Drosophila melanogaster populations that had undergone over 400 generations of artificial selection on body size to investigate the changes in life history traits associated with the evolution of extremely large and extremely small body sizes. Populations selected for small body size experienced strong trade-offs in multiple life history traits, including reduced female fecundity and lower juvenile viability. Although we found correlated changes in egg size associated with selection for both large and small body size, after adjusting for female body size, females from populations selected for large size had the lowest relative investment per egg and females from populations selected for small size had the highest relative investment per egg. Taken together, our results suggest that egg size may be a key constraint on the evolution of body size in D. melanogaster, providing insight into the broader phenomenon of body size evolution in insects.

README: Data from: Life history changes associated with over 400 generations of artificial selection on body size in Drosophila

https://doi.org/10.5061/dryad.15dv41p4v

Description of the data and file structure

BodySizes.csv
Male and female body size data for the Body Size experiment. Data are displayed in Figure 1.
“Block” = the experimental block
“Treatment” = the artificial selection treatment (L, S, C, or LHm)
“Lineage” = the artificial selection lineage (e.g., L1, L2, S1, etc.)
“SieveHoleDiameter” = sieve hole diameter (in 𝜇m) of the sieves used for size sorting
“NumberMales” = the total number of males found within each sieve after size sorting
“NumberFemales” = the total number of females found within each sieve after size sorting

Fecundity_Lin.csv
Fecundity data for the lineage comparisons (includes LHm). Data are displayed in Figure 2b.
“Lineage” = the artificial selection lineage (e.g., L1, L2, S1, etc.)
“Treatment” = the artificial selection treatment (L, S, C, or LHm)
“Vial” = the experimental vial that 10 females were sampled from
“Female” = each female sampled from “Vial”
“Fecundity” = the total number of eggs laid by the female

Fecundity_TM.csv
Fecundity data for the treatment comparisons (does not include LHm). Data are displayed in Figure 2a.
“Treatment” = the artificial selection treatment (L, S, or C)
“Lineage” = the artificial selection lineage (e.g., L1, L2, S1, etc.)
“Vial” = the experimental vial that 10 females were sampled from
“Female” = each female sampled from “Vial”
“Fecundity” = the total number of eggs laid by the female

EggVolumeImages.zip
Digital images of eggs measured for the Egg Size experiment. Each folder contains the images for a given experimental block (i.e., B1 = Block 1). The JPEG file names use the following format: Block-Lineage-Image # (e.g., B1-C1-001 is the first image of C1 eggs from Block 1). Within a block, each lineage has 1-3 replicate images, as needed, to ensure 10 representative eggs on their dorsal sides were available to measure. All images were taken at 40X magnification. For additional precision, the folder for each block also contains an image of 1 mm for scale. The scale image for Block 6 has unfortunately been lost, but this block was conducted identically to the previous blocks.

EggVolumes.csv
Egg volume data for the Egg Size experiment. Data are displayed in Figure 3.
“Block” = the experimental block
“Treatment” = the artificial selection treatment (L, S, C, or LHm)
“Lineage” = the artificial selection lineage (e.g., L1, L2, S1, etc.)
“EggVolume” = egg volume (in 𝜇m^3)

RelativeEggInvestment.csv
Relative egg investment data. Data are displayed in Figure 4.
“Lineage” = the artificial selection lineage (e.g., L1, L2, S1, etc.)
“Treatment” = the artificial selection treatment (L, S, C, or LHm)
“RelativeEggInvestment” = relative egg investment, calculated using the mean female body size and mean egg size for each lineage

Viability_Lin.csv
Viability data for the lineage comparisons (includes LHm). Data are displayed in Figure 5b.
“Block” = the experimental block
“Treatment” = the artificial selection treatment (L, S, C, or LHm)
“Lineage” = the artificial selection lineage (e.g., L1, L2, S1, etc.)
“OLRE” = an observation-level random effect added to potentially reduce overdispersion
“NumHatch” = the number of eggs out of 24 that hatched
“NumNotHatch” = the number of eggs out of 24 that did not hatch
“NumEclose” = the total number of eclosing adult flies
“NumEggNotEclose” = the total number eggs (out of 24) that did not eclose as adult flies
“NumLarvaeNotEclose” = the total number of larvae (from “NumHatch”) that did not eclose as adult flies

Viability_TM.csv
Viability data for the treatment comparisons (do not include LHm). Data are displayed in Figure 5a.
“Block” = the experimental block
“Treatment” = the artificial selection treatment (L, S, C, or LHm)
“Lineage” = the artificial selection lineage (e.g., L1, L2, S1, etc.)
“OLRE” = an observation-level random effect added to potentially reduce overdispersion
“NumHatch” = the number of eggs out of 24 that hatched
“NumNotHatch” = the number of eggs out of 24 that did not hatch
“NumEclose” = the total number of eclosing adult flies
“NumEggNotEclose” = the total number eggs (out of 24) that did not eclose as adult flies
“NumLarvaeNotEclose” = the total number of larvae (from “NumHatch”) that did not eclose as adult flies

Sharing/Access information

Links to other publicly accessible locations of the data: N/A

Data was derived from the following sources: N/A

Code/Software

JMP logs for body size, egg volume, and relative egg investment analysis

All analysis performed using JMP 17.

BodySizesJMPlogs.rtf
Output logs containing the JMP scripts for the following analyses using the body size data in BodySizes.csv:

the ANOVA testing for an effect of artificial selection Treatment on Female Body Size (Table S2a)
the ANOVA testing for an effect of artificial selection Treatment on Male Body Size (Table S2b)
the ANOVA testing for an effect of artificial selection Lineage on Female Body Size (Table S3a)
the ANOVA testing for an effect of artificial selection Lineage on Male Body Size (Table S3b)

EggVolumesJMPlogs.rtf
Output logs containing the JMP scripts for the following analyses using the egg size data in EggVolumes.csv:

the ANOVA testing for an effect of artificial selection Treatment on Egg Volume (Table S6a)
the ANOVA testing for an effect of artificial selection Lineage on Egg Volume (Table S7)

RelativeEggInvestmentJMPlog.rtf
Output log containing the JMP script for the ANOVA testing for an effect of artificial selection Treatment on Relative Egg Investment (Table S6b).

R scripts for GLMMs used to analyze female fecundity and juvenile viability data.

All analyses performed using R 4.3.1 and the following packages:
glmmTMB 1.1.7
Car 3.1-2
emmeans 1.8.7

Fecundity.R
R code for the following analyses:

the negative binomial GLMM used for lineage comparisons of the fecundity data in Fecundity_Lin.csv (Table S5)
the negative binomial GLMM used for treatment comparisons of the fecundity data in Fecundity_TM.csv (Table S4)

Viability.R
R code for the following analyses:

Egg-to-larva viability (hatch rate)

binomial GLMM used for lineage comparisons of the viability data in Viability_Lin.csv (Table S9b)
binomial GLMM used for treatment comparisons of the viability data in Viability_TM.csv (Table S8b)

Larva-to-adult viability

binomial GLMM used for lineage comparisons of the viability data in Viability_Lin.csv (Table S9c)
binomial GLMM used for treatment comparisons of the viability data in Viability_TM.csv (Table S8c)

Egg-to-adult viability

binomial GLMM used for lineage comparisons of the viability data in Viability_Lin.csv (Table S9a)
binomial GLMM used for treatment comparisons of the viability data in Viability_TM.csv (Table S8a)

Data from: Life history changes associated with over 400 generations of artificial selection on body size in Drosophila

Data files