1. Many insects that live in temperate zones spend the cold season in a state of dormancy, referred to as diapause. As the insect must rely on resources that were gathered before entering diapause, keeping a low metabolic rate is of utmost importance. Organs that are metabolically expensive to maintain, such as the brain, can therefore become a liability to survival if they are too large.

2. Insects that go through diapause as adults generally do so before entering the season of reproduction. This order of events introduces a conflict between maintaining low metabolism during dormancy, and emerging afterwards with highly developed sensory systems that improve fitness during the mating season.

3. We investigated the timing of when investments into the olfactory system are made by measuring the volumes of primary and secondary olfactory neuropils in the brain as they fluctuate in size throughout the extended diapause life-period of adult Polygonia c-album butterflies.

4. Relative volumes of both olfactory neuropils increase significantly during early adult development, indicating the importance of olfaction to this species, but still remain considerably smaller than those of non-diapausing conspecifics. However, despite butterflies being kept under the same conditions as before the dormancy, their olfactory neuropil volumes decreased significantly during the post-dormancy period.

5. The opposing directions of change in relative neuropil volumes before and after diapause dormancy indicate that the investment strategies governing structural plasticity during the two life-stages could be functionally distinct. As butterflies were kept in stimulus-poor conditions, we find it likely that investments into these brain regions rely on experience-expectant processes before diapause and experience-dependent processes after diapause conditions are broken.

6. As the shift in investment strategies coincide with a hard shift from pre-mating season to mating season, we argue that these developmental characteristics could be adaptations that mitigate the trade-off between dormancy survival and reproductive fitness.

Species and rearing

The Comma butterfly Polygonia c-album (Linné, 1758) has a large distribution area, spanning from Great Britain and northern Africa in the west, to Japan in the east. It is a facultatively multivoltine species, having more than one generation per year in areas where the summer season is sufficiently long, and undergo diapause as imago. Uniform daylength, especially if shorter than 13 hours, favours the developmental path leading to diapause readiness, while increasingly longer days (e.g. L:D 12:12 --> 22:2) experienced during larval development favours the path leading to directly reproducing adults, especially when accompanied with increasingly higher temperatures (e.g. 17°C --> 23°C).

P. c-album  is considered a polyspecialist with regards to host plant utilisation, and olfaction is likely to play an important role during the search for suitable oviposition sites as mated females are known to show attraction towards host plants based solely on olfactory cues (Schäpers et al. 2015).

A cohort of P. c-album females, derived from a lab population originating in the Norfolk region of the United Kingdom and provided by the company World Wide Butterflies, was used in this study. Larvae from approximately 30 egg-laying females were reared on fresh cuts of stinging nettles Urtica dioica under a light and temperature regime promoting the development of diapausing adults (12:12 L:D, 17°C). Pupae were removed from their pupation sites two days after pupation, sexed by inspection of the genital slits, and placed in paper lined cups covered with mesh nets until adult eclosion.

Conditions during treatment

One day old adults were transferred to mesh net cages (50cm*50cm*50cm) in a secluded room with a neutral olfactory environment. The room was under a 6:18 L:D light cycle, and a dynamic temperature cycle peaking at 27°C during the end of the light period, and dropping down to 20°C during the dark period. Cages were equipped with feeding stations consisting of a plastic cup with a dish sponge and freshly mixed 25% sugar water. Feeding stations were regularly exchanged in order to prevent growth of algae and fermentation of the sugar as to avoid non-standardized variations in gustatory and olfactory conditions during the time spent in flight cages.

Butterflies were kept in the flight cages for 14 days before the diapause dormancy climate conditions were initiated, and again for two weeks after the dormancy conditions were terminated. Throughout the period when dormancy conditions were active, the butterflies were kept in plastic cups covered with mesh nets (secured with rubber bands) and placed in a cardboard box which in turn was kept in a sheltered area on the roof of the Zoology Department of Stockholm University during the winter of 2018-2019.

All butterflies were placed on the underside of the mesh net, inside the cups, as this position seems to increase diapause survival rates (Christer Wiklund, Stockholm University, 2018.07, personal communication). The animals were regularly checked on during the dormancy period and individuals which had dropped off from the net were gently put back into their original position if still alive at the time of discovery.

Sampling

Butterflies were sampled at five different time points: 1) 1 day after adult eclosion, 2) immediately before the onset of dormancy conditions, at 14 days after adult eclosion, 3) during the mid-stage of the diapause dormancy period, approximately two months after dormancy conditions were initiated, 4) at the end of diapause dormancy, approximately four months after dormancy conditions were initiated, and 5) two weeks after dormancy conditions were terminated and butterflies had been returned to indoor flight cages.

It should be noted that P. c-album butterflies generally take flight within minutes and start feeding within hours after diapause dormancy climate conditions are terminated, and are ready to mate within a few days up to one week (M. E., unpublished observation).

Dissection and sample preparation

Butterflies were decapitated using micro scissors, and their heads were fixated overnight at 4°C in a 4% paraformaldehyde (Sigma-Aldrich, Steinheim, Germany) solution. Heads were washed in phosphate buffered saline (Sigma-Aldrich, Steinheim, Germany) containing 2,5% Triton-X (Sigma-Aldrich, Steinheim, Germany) (PBStx) 4x15 min before the brains were dissected out. Brain samples were incubated in a 1:20 solution of primary antibody targeting synapsin (3C11 anti-SYNORF1, Developmental Studies Hybridoma Bank, Iowa City, IA, USA) and PBStx with 0,5% bovine serum albumin (BSA) (Sigma-Aldrich, St. Louis, MO, USA) over 3 days at 4°C on a shaker at low setting. Samples were then washed 5x1h before incubation in secondary antibody (AlexaFluor 488 (Life technologies, Eugene, OR, USA) 1:500 in PBStx 0,5% BSA) for 3 days at the same conditions as for primary incubation. Stained samples were washed 4x1h in PBStx, 1x1h in PBS, and treated for optical clearing in Omnipaque (GE Healthcare AS, Oslo, Norge), firstly in a 1:1 solution of Omnipaque:PBS for 24h and then stored in pure Omnipaque for a minimum of 24h. As clearing in Omnipaque eliminates the need for sample dehydration in ethanol and usage of methyl salicylate, the tissue shrinking associated with such protocols (Bucher et al. 2000; Smolla et al. 2014) is avoided, allowing for more accurate volume measurements. Omnipaque is an odourless and benign liquid commonly used in clinical treatments, making it safe and easy to work with.

Confocal scanning and reconstruction

Cleared samples were whole-mounted in omnipaque on glass slides with custom-made 0,5 mm metal spacers and were optically sectioned using a Zeiss LSM 780 META (Zeiss, Jena, Germany) scanning confocal laser microscope. Images with a resolution of 1024 x 1024 pixels were obtained with a 10x air objective. Each section had a thickness of approximately 3 mm, resulting in image stacks of about 100 sections per sample.

3D reconstruction and extraction of volumetric measurements (in mm³) from antennal lobes, mushroom body calyces, and whole central brain regions (optic lobes were excluded from analysis) were performed by using the native segmentation, volume rendering, and surface reconstruction tools in the Thermo Scientific^TM AMIRA^TM (v. 2019.3) image processing software.

Samples that were physically damaged during technical processing were excluded from analysis, as were those that did not render high quality images during confocal scanning, or had physical abnormalities such as divergent number of calycal cups. In cases where only one of a paired neuropil was discarded for the previously stated reasons, the intact neuropil volume was duplicated as to achieve total neuropil volumes comparable to fully intact samples. As paired neuropils are known to be symmetrical between hemispheres (Galizia et al. 1998; el Jundi et al. 2009; Eriksson et al. 2019) both neuropils were discarded in cases where the volume difference was greater than 10% within a pair. The latter phenomenon is to be expected in cases where the neurolemma was ruptured before samples were fully fixated, allowing damaged tissues to expand post mortem.