Meiotic drive adaptive testes enlargement during early development in the stalk-eyed fly.

The sex ratio 'SR' X-linked meiotic drive system in stalk-eyed flies destroys Y-bearing sperm. Unlike other SR systems, drive males do not suffer fertility loss. They have greatly enlarged testes which compensate for gamete killing. We predicted that enlarged testes arise from extended development with resources re-allocated from the accessory glands, as these tend to be smaller in drive males. To test this, we tracked the growth of the testes and accessory glands of wild-type and drive males over 5–6 weeks post-eclosion before males attained sexual maturity. Neither of the original predictions is supported by these data. Instead, we found that the drive male testes were enlarged at eclosion, reflecting a greater allocation of resources to the testes during pupation. Testes grow at a higher rate during early adult development, but there was no evidence that this retards the growth of the accessory glands. Further experiments are proposed to investigate whether smaller accessory glands only arise in drive males post-copulation or when flies are subjected to nutritional stress. Our experimental findings support the idea that enlarged testes in drive males arise as an adaptive allocation of resources to traits that enhance male reproductive success.Microsoft ExcelFunding provided by: Department of Genetics, Evolution, and Environment*Crossref Funder Registry ID: Award Number: NE/S007229/1Funding provided by: Engineering and Physical Sciences Research CouncilCrossref Funder Registry ID: http://dx.doi.org/10.13039/501100000266Award Number: EP/F500351/1Funding provided by: Natural Environment Research CouncilCrossref Funder Registry ID: http://dx.doi.org/10.13039/501100000270Award Number: NE/R010579/1Funding provided by: Biotechnology and Biological Sciences Research CouncilCrossref Funder Registry ID: http://dx.doi.org/10.13039/501100000268Award Number: BB/V003542/1Funding provided by: Engineering and Physical Sciences Research CouncilCrossref Funder Registry ID: http://dx.doi.org/10.13039/501100000266Award Number: EP/I017909/1Flies were dissected on days 0, 1, 4, 8, 12, 16, 20, 34 and 56 (N = 24–53 per time point, electronic supplementary material, tables S1 and S2). A follow-up experiment was carried out with more intense measurements from day 11 to day 25 (short dataset), with dissections on days 11, 13, 15, 17, 19, 21, 23 and 25 (N = 37–53 per time point, electronic supplementary material, tables S3 and S4). The thorax and eyespan of ice-anaesthetized flies were measured prior to dissection, using an Infinity Capture video microscope attached to a computer equipped with NIH image software (FIJI (Image J) Version 2.1.0/1.53c). The thorax was measured from the prothorax anterior tip, along the midline ending at the joint in-between the thorax and metathoracic legs [36]. Eyespan was measured from the outer tips of the eyes adjacent to where the stalk joins the eye bulb [38]. Flies were then dissected in 15 µl of phosphate-buffered saline (PBS) using 5 mm forceps on a glass slide under the stereomicroscope. The testes and accessory glands were isolated then untangled and uncoiled without causing rupture or damage. Excess material such as external cuticle was removed from the slide to prevent distortion of the image. Another 15 µl of PBS was added before adding a glass cover. Images were taken using a differential interference contrast microscope on QCapture Pro imaging software at either x5 or x10 magnification. The polygon selection tool in Image J was used to take area measurements for both the testes and accessory glands, by tracing around the outline of the organs. All statistical analyses were performed in R (v.1.4.1103). Linear regression models were used to identify differences in reproductive organ size between genotypes. Models included genotype, age (days), thorax size (body size) and relative eyespan. A stepwise build-up was used to add terms that improved the model fit. Terms that did not improve the model fit were discarded. The morphological traits of thorax size and relative eyespan were added as covariates as they are known to differ between genotypes, and correlate with reproductive organ size in mature adult flies [34]. Relative eyespan was calculated from the residuals using a linear regression model after taking into account thorax size, as these traits are strongly collinear [39]. To determine trade-offs between the development of the testes and accessory glands with other morphological traits, interaction terms were tested. Specifically, break-point analysis was conducted to investigate the interaction between genotype and age (days) on the testes as a proxy of growth rate. Mean and standard error trait sizes (mm) are reported throughout. See electronic supplementary material, for all models