Data from: Response to MHC-based olfactory cues in a mate choice context in two species of darter (Percidae: Etheostoma)

Animal care and use statement This research was conducted with the approval and oversight of the Bloomington Institutional Animal Care and Use Committee (BIACUC protocol numbers 19-003, 17-016, and 20-015). Scientific collection permits were issued to KMM by the Indiana Department of Natural Resources. 2.1 Subject collection Individuals of the rainbow darter (Etheostoma caeruleum) and the fantail darter (Etheostoma flabellare) were collected with a seine net from Clear Creek, Bloomington, Indiana, Monroe County (39.12118432195659, -86.53886585996088) on April 6, 2019. We identified individuals to species based on morphological characteristics (Page and Burr, 2011; Simon, 2011). Individuals were sexed based on breeding characteristics. In both species, we identified females as gravid by the presence of distended abdomens. We identified male rainbow darters in breeding condition by nuptial coloration, and male fantail darters in breeding condition by darkened body color and the presence of egg mimics on the first dorsal fin (Kuehne and Barbour, 1983; Page, 1983). Individuals were transported to Indiana University in aerated, covered buckets in an air-conditioned vehicle. 2.2 Care and housing Individuals were housed separately by sex and species. Males and females of the same species were housed such that they were unable to see conspecifics of the opposite sex. We housed individuals in 20-gallon tanks (30" x 12" x 12") equipped with power filters and air pumps mounted with large air stones. River rock aquarium gravel was provided as substrate. Small terra cotta pots were provided as hides. We conditioned the tank water with Amquel Plus and performed 20% water changes once per week. Individuals were fed frozen brine shrimp and bloodworms once per day, as much as could be consumed in 5 minutes. 2.3 Fin clipping and VIE tagging Individuals were fasted overnight ahead of the fin clipping and tagging procedures. Individuals were anesthetized with MS-222 (25 mg/L) until gill movement was slow and steady and the subject did not respond to touch. Paper towels moistened with conditioned water were placed on the gill covers to maintain the subject’s breathing during the procedure. Each subject was measured for standard length (tip of the nose to the caudal peduncle). A small (3mm2) piece of fin tissue was removed from the anal fin of each subject. Iodine was used to sterilize the fin clip site during the procedure. Each individual was injected with three visible implant elastomer (VIE) tags (Northwest Marine Technology), one anterior to the first dorsal fin, one between the first and second dorsal fin, and one behind the second dorsal fin. Each VIE tag combination was a unique color pattern used to identify the individual. The total procedure from start to finish took approximately three minutes. Individuals were place in conditioned water and allowed to recover from the anesthesia. Recovery was pronounced when the subject resumed normal swimming and breathing. All individuals were monitored over the following 24 hours for full recovery. All fin tissue regenerated within 3 weeks, and all VIE tag wounds healed within one week with no sign of infection or long-term effects. Behavioral trials were not performed until all individuals had fully healed and recovered from fin clipping and VIE tagging. Fin clips were frozen and stored for genotyping. 2.4 MHC genotyping All individuals were genotyped at a 164 base pair portion of the MHC class IIb gene (exon 2) according to deep-sequencing methods developed for fish (Lighten et al., 2014). We used custom primers and protocols we had developed for previous research on darters (Million and Lively, 2022). Our prior research indicated the presence of up to 5 MHC IIb loci in both our species of interest, meaning an individual genotype could contain 1 to 10 alleles. The number of alleles was recorded for each genotype as well as the identity of each allele in each genotype. 2.5 Breeding conditions in the lab To maintain the breeding condition of all individuals in the lab during the trials, we manipulated ambient light and temperature to mimic the conditions of the natural breeding season. We kept the day length cycle at 14L:10D and maintained the ambient room temperature at 28 degrees Celsius. We assessed the success of our efforts by observing female gravidity and male nuptial characteristics as described above. 2.6 Construction of motorized visual stimuli To control for possible differences in visual and behavioral cues between live males, we constructed motorized visual stimuli with which to pair the olfactory cues during the trials. These models allowed us to present identical visual stimuli to the focal females and provide a visual context in which to interpret the olfactory stimuli. Motorized, painted models have been successfully used by other researchers for dichotomous choice trials in darters (Williams, Gumm, and Mendelson, 2013, Williams and Mendelson, 2010). 3D models of a male fantail darter and a male rainbow darter were purchased from the website TurboSquid (model artist: “adicaza”) (Figure 1). Files were modified for compatibility with a 3D printer. Life-sized male models were 3D printed using a Formlabs Form 2 printer and Formlabs SG dental resin. 3D printing was performed in the Lennon lab at Indiana University. High-resolution images of a male fantail darter and a male rainbow darter were obtained from the North American Native Fishes Association photo gallery, and color palettes were developed from colors isolated from the images. Models were painted with colors and patterns resembling a male of each species in breeding condition (Figure 2). Models of the same species were painted to be visually identical. Small glass amber-colored beads (2 mm outside diameter) were glued to the first dorsal fin of the fantail darter models to resemble egg mimics. Models were mounted onto stepper motors using magnets and transparent plexiglass platforms. Stepper motors were attached to microcontrollers, which in turn were connected to an Arduino Uno REV3 unit. Model movements were programmed using the software Arduino IDE 1. The Arduino unit was programmed so that the models made synchronized movements that resemble the courtship displays of live breeding males. Two visually identical models of each species were created. 2.7 Experimental setup A 37.9-liter tank (50.8 cm" x 25.4 cm x 30.5) was set up in a location where the subject would have no visual contact with other individuals during the trial (Figure 1). The focal tank was surrounded with sheets of panda paper (Vivasun) to block external light and visual stimuli. The bottom of the tank was blacked out using matte black aquarium decal material. The tank was separated into three zones of preference, marked by narrow white lines. The neutral zone was centered in the middle and took up 80% of the tank bottom surface area. Each preference zone on the left and right took up 10% of the tank bottom surface area. Ambient temperature in the trial area matched that of the housing area. The tank was lit using ambient fluorescent light and a Nicrew ClassicLED Plus full-spectrum aquarium light mounted above the tank. Two 10-gallon tanks were placed underneath the focal tank, completely out of visual range of the focal subject. All tanks were partially filled with conditioned, aerated water (3/4 full for the focal tank, ½ full for each olfactory stimulus tank). At the conclusion of each trial, all tanks were emptied, scrubbed with 95% ethanol, allowed to air dry, and refilled with the same volume of fresh water. An AKASO EK7000 Pro 4K Action Camera was mounted near the tank to record all trials. All motorized and electronic devices were connected to a remote control so that they could be turned on and off without the observer in the room during a trial. The trials were monitored outside the room in real time using a cell phone remotely connected to the camera. We pre-determined that if a participant showed signs of severe stress (gasping, attempting to jump out of the tank) at any point during a trial, then the trial would have been immediately terminated and the subject returned to housing. However, during these trials no individuals displayed signs of extreme stress, so we did not terminate any of the trials. No individual participated in more than one trial per 24 hour period, either as a subject or as a stimulus individual. Submersible pumps (Homasy 300 L/H, 4 W) were placed into each olfactory stimulus tank. The pumps were connected to the focal tank via clear food grade vinyl tubing (5/16 ‘’ ID, 7/16 ‘’ OD). The pumps were primed with clean conditioned water before and after every trial. Plastic clamps were used to control the water flow and to ensure that the water flowed into the focal tank at the same rate on each side. This was visually confirmed prior to the start of every trial. 2.8 MHC-based choice trials We used dichotomous choice trials to assess the response of females of each species to MHC-based olfactory cues from conspecific males (Figure 3). The method is standard for evaluating mate choice preferences in darters (Mattson et al., 2020; Williams et al., 2013). For each trial, to control for visual and behavioral differences between the two live males, two visually identical male models of the same species as the focal female were placed on either side of the focal tank and mounted on stepper motors. Before the start of the trial, blinds were used to block the view outside the tank of the focal female during acclimation. One focal female was acclimated to the focal tank for a minimum of 10 minutes. A female was considered acclimated when she initiated normal free swimming around all sides of the tank and exhibited normal breathing patterns. One live stimulus male was acclimated to each of the olfactory stimulus tanks out of view of the focal female for 10 minutes, the same amount of time that the focal female was acclimated to the focal tank. After the acclimation period, the blinds were removed, and the focal female was allowed to explore the tank and visually inspect each model. The trial began after the female had crossed into both preference zones, seen both models, and returned to the neutral zone. A trial did not progress if the focal individual did not cross both sides and return to the middle. At the start of the trial, both male models were turned on to provide identical visual displays for the female, and water was pumped into the focal tank on either side from each of the stimulus tanks. During each trial the focal female was presented with the scent of a male with an MHC genotype similar (differing by no more than one allele) or identical to her own on one side, and one male with an MHC genotype completely dissimilar to her own (none of his alleles matched hers) on the other side. To control for heterozygosity, we matched stimulus males so that they either had the same number of alleles in their genotypes or only differed in number by one. Due to small sample sizes, some females participated in more than one trial. In these cases, a female was always presented with a completely novel pair of stimulus males. The trial period lasted for 10 minutes, and the entire trial was observed outside the room by camera and was recorded. To control for potential side bias in the focal females, we switched the sides on which the similar and dissimilar male scents were presented to the focal female between each trial, and we switched the sides of the two models between each trial. Sample sizes for the trials were N = 4 for rainbow darters and N = 5 for fantail darters. Total number of trials run were N = 9 for rainbow darters and N = 2.9 Trial analyses Videos were analyzed using the software Boris (Friard and Gamba, 2016). We constructed an ethogram in the software to use during analysis. All videos were scored for the amount of time the focal female spent in each of the three zones of the focal tank. If during a trial a female spent more than 80% of her time in the neutral zone, we concluded that she had not interacted enough with the stimuli for us to score her preferences, and that trial was set aside from further analysis. A strength of preference (SOP) equation was used to evaluate the responses to the stimuli. The equation is SOP = (S – D)/(S + D), where S = the time spent in the preference zone with the male with a similar MHC genotype and D = the time spent in the preference zone with the male with a dissimilar MHC genotype. The values obtained range from a minimum of -1 (strong preference for dissimilar genotype) to a maximum of 1 (strong preference for a similar genotype). A value at or near 0 indicates no strong preference for either genotype. Further data analysis was conducted using the software R version 4.1.0 (R Development Core Team, 2018). Mean SOP between the species was compared using a Wilcoxon rank-sum test. Proportion of time spent with the male with the similar MHC genotype (as opposed to the male with the dissimilar MHC genotype) was tested against the null hypothesis that the females spent equal time with both males using a one-sample Wilcoxon test. Power analyses on the data were conducted using the software G*Power.