Summary
Female choice for male ornamental colouration has been demonstrated in a number of different taxa. Among fishes, most studies have been conducted in a laboratory setting and show that females prefer more colourful male ornaments. In this study, we observed female bluegills (Lepomis macrochirus) spawning in their natural environment and compared spawning behaviours to male traits and position within a colony. We observed spawning activities of 76 parental males in Lake Opinicon, Ontario. We captured each male and used reflectance spectrometry to objectively quantify the colour of six body regions and measured morphological characteristics. Our results show that female spawning behaviours did not significantly differ between central and peripheral males, although egg scores were higher in central nests. During spawning, females appeared to enter the nests of parental males haphazardly. However, our results suggest that male cheek colouration influenced the number of females spawning, the number of eggs they released, and the amount of time they spent in the nest. Moreover, male breast colouration significantly predicted reproductive success as quantified through egg scores. Together, our findings suggest that females may use male cheek and breast colouration, condition-dependent sexual ornaments, as key traits on which to base their mate choice decisions.
Methodology
Spawning observations Between May 31 and June 27, 2008, we observed bluegill spawning activities at three colonies of similar depth (1.4-1.7 m) in Lake Opinicon, ON, Canada (44°34'N, 76°19'W). At the beginning of the season, we monitored previously used nesting colonies (indicated by clusters of saucer-like depressions in the substrate) throughout the lake for the arrival of parental males. If spawning within a colony commenced, we stayed and observed this colony for the day. When we arrived at a spawning colony, we marked each active nest with a numbered tile for identification. We defined an active nest as one that already contained eggs and where the parental male was actively court ing and guarding. This distinction is important as the females may prefer nests where eggs are already present (e.g., Forsgren et al., 1996 and references within). We observed each nest within the colony for 30 min, 2-3 times throughout the day (morning and afternoon) to account for any possible diel variation in activity, for a total of 60-90 min of observations per male. Observation times were converted into a 5-point scale, where each point represented a range of two hours, starting from 0800 and ending at 1800 h. During the 30-min observation period, each observer was responsible for documenting the activities of 1-4 nests within the colony. Previously, Colgan et al. (1979) showed that the presence of an observer does not cause large disturbances in the spawning activities of bluegills.
Observations were made by 2-3 snorkelers and were supplemented with an underwater video camera (Sony® Handycam HDR-SR8, 100GB HDD housed in Amphico® DiveBuddy evoHD elite). The video unit was positioned on a tripod to record activities at multiple nests, and video recordings were analyzed by two separate observers. For each nest, we recorded male location (central or peripheral) in the colony, number of females that entered the nest area, number of females that spawned with the parental male, number of tilts (release of approx. 10-30 eggs; Neff et al., 2003) performed by each female, duration of the spawning event, and the number of aggressive behaviours performed by the parental male toward actual or potential intruders (other parental males and sneaker males). We defined central males as those that had at least one nest between themselves and the edge of the colony, whereas peripheral males had at least one edge of the nest exposed and adjacent to no other nest (Gross & MacMillan, 1981). We converted each observation measurement to a rate in minutes based on the total observation time (typically 60-90 min). Other studies have investigated mate choice in lekking species using similar methods to those described here (e.g., McKaye et al., 1990; Young et al., 2009).