Effects of mucus trail following on the distance between individuals of opposite sex and its influence on the evolution of the trait in the Ezo abalone Haliotis discus hannai

Study species

Haliotis discus species is a marine gastropod mollusk in the family Haliotidae, which is distributed in the waters off Japan and eastern Asia. The species mainly inhabits rocky shores. The spawning season in the Iwate Prefecture, Japan, is from August to late October, when water temperatures are around 20–21 °C after the effective accumulated temperature (EAT; Uki & Kikuchi, 1984) reaches 1000 degree-days. Experimental animal subjects were captured in the ocean of Iwate Prefecture in October 2015 and segregated by gender in tanks equipped with water temperature control.

Mucus trail following

To evaluate whether H. discus hannai follows mucus trails, binary choice tests were conducted using a y-maze, taking place on December 11–28, 2015, August 17–20, 2016, and June 26-July 17, 2017. In this study, individuals with immature gonads were classed as non-reproductive individuals (experiments conducted during December 11–28, 2015) and individuals at simulated EAT 1000–2000 degree-days were classed as reproductive individuals (experiments conducted during August 17–20, 2016, and June 26–July 17, 2017). The y-maze consisted of a 50-cm stem and 50-cm arms (width, nine cm; height, five cm). Further details of the experimental abalones are presented in Data S1. The experiments were conducted as follows.

Step 1: To form the mucus trail on the surface of the y-maze, one arm of the y-maze was closed using a plastic plate, thus restricting where marker individuals were allowed to move (right arm: n = 38, left arm: n = 37). The marker individual was released at the end of the stem in the daytime and usually began moving within one hour, eventually reaching the end of the y-maze arm after which is was removed. The plastic plate was then removed to open both arms of the y-maze. If the marker individual remained at the end of the stem for over a day they were replaced with a new marker individual.

Step 2: The subject individual was released at the end of the stem and its direction of creeping was observed. This step commenced at 18:00 h because H. discus hannai individuals are active after sunset. Followers often reached the end of the arms within several minutes. If a follower remained at the end of the stem for over 30 min, its direction of creeping was observed using a video camera (DVSA10FHDIR; Kenko Tokina Corporation, Tokyo, Japan) until 6:00 h. If a follower individual remained at the end of the stem for over a day, the experiment was stopped. After finishing each experiment, the walls and floor of the y-maze were scrubbed clean using a spongy, prior to beginning subsequent experiments.

The effects of marker-tracker combinations and season on MTF rates were tested using a generalized linear model (binomial distribution) with likelihood ratio tests using the “Anova” function in the “car” package for R v3.5.2. We also used a binomial test to determine whether MTF was actually being observed, without dividing the subjects into experimental groups.

Mucus trail following as a mate-finding strategy

To examine whether MTF behavior reduces the distance between members of the opposite sex, an IBM consistent with the behavior of H. discus hannai was constructed (see also Fig. S1) using NetLogo v5.3.1. The script is available on GitHub with the following file name: File S1 (https://github.com/YMatsumoto5536/PeerJ-netlogo).

In this model, female MTF abalones (i.e., mutant females), and non-MTF female and male abalones (wild females and wild males) were generated. The IBM compared the distances between wild males and mutant females and between the wild males and wild females. The comparison revealed whether MTF is effective for reducing the distances between the members of the opposite sex from the aspect of female benefit.

Movement algorithm

Mutant females searched the grid for a mucus trail within a 10 cm search radius. They next decide whether to move to a grid cell with mucus or an empty grid cell according to their individual mucus following rate parameter (see below). The individual moving distance for each day was generated from an exponential distribution with mean of 48.76 cm, which is consistent with observed behavior in the holding tanks (Data S2). The individual MTF rate was generated from a binomial distribution (trial number = 75, number of arms with mucus = 60, estimated following rate = 0.8). This following rate is consistent with the observed behavior in the tank experiments (Fig. 1). Model abalones followed mucus trails irrespective of the sex combination, consistent with the only small sex combination effects observed in the tank experiment (Fig. 1). Although the wild females and wild males moved the same distance as the mutant females, their direction of movement was random. All individuals left mucus trails before moving to the next grid cell. Studies have suggested that the mucus trail of gastropods is degraded by bacteria within a single day (Herndl & Peduzzi, 1989; Peduzzi & Herndl, 1991). Therefore, we set the functional persistence period of the mucus to one day in the IBM. These processes were repeated until the day’s determined locomotion distance had been reached. At the end of the 30 day test period the distances between opposite sex individual was measured.

Evolutionary model of MTF behavior

To examine whether abalones evolved the ability to follow mucus trails to reduce distances between members of the opposite sex, we constructed simple population genetic models for a diploid population undergoing nonoverlapping, discrete generations (see also Fig. S2). It should be noted, however, that H. discus hannai undergoes repeated spawning over several years. The script is available on GitHub with the following file name: File S2 (https://github.com/YMatsumoto5536/PeerJ-netlogo).

Following the mucus trail of another individual

The MTF rates observed in each experiment are shown in Fig. 1. The marker and follower combinations (likelihood-ratio test, df = 4, χ² = 2.68, p = 0.61), season (df = 1, χ² = 0.026, p = 0.87), and their interaction term (df = 3, χ² = 0.36, p = 0.95) did not affect the MTF rate. Abalones chose the arm with a mucus trail more frequently (n = 60) than the arm without a mucus trail (n = 15) when experimental groups were pooled (binomial test, p = 1.588 e−07).

MTF reduces the distance between opposite sex individuals

The distance between wild males and mutant females (mean ± SD = 10.83 ± 6.57 cm ) was significantly less than that between wild males and wild females (11.08 ± 6.61 cm; df = 1, F = 6.90, p = 0.009); however, the estimation difference in distance was only 2.4 (±SD = 0.8) cm.

Evolutionary role of reducing opposite-sex distances in MTF

As the frequency of D alleles did not increase over the generations in the control model (Fig. 2A), the increasing frequency of D alleles in the trail following model was caused by an increased fertilization rate rather than random genetic drift. In the MTF model, the frequency of D alleles (Fig. 2B, see also Data S3) increased over the generations 53 of the 100 simulations. Complete disappearance of the D allele occurred within 314 generations (average ± SD = 88 ± 69 generations; range = 13–314) in the MTF model.

The MTF trait

Our experiments strongly suggest that the presence of a non-self mucus trail affects the directional decision making in H. discus hannai. Mucus trails of H. discus hannai are known to attract their larvae (Roberts, 2001); thus, it is not surprising that adult individuals also react to conspecific mucus trails. The tentacles of H. asinina react to protein extracted from mucus trails in vitro (Kuanpradit et al., 2012), further supporting the hypothesis that H. discus hannai senses mucus trails using their tentacles.

We found that H. discus hannai followed mucus trails irrespective of the marker’s sex or season. As the experimental design did not allow for the selection of mucus from different sexes simultaneously, we cannot deny the possibility that H. discus hannai might prefer mucus from the opposite sex and sex-related differences in preference for mucus, although sex-related differences in mucus composition have not been found in H. asinina (Kuanpradit et al., 2012). Although the preference for mucus from the opposite sex in H. discus hannai was not detected in this study, there remains the possibility that mucus-trail-following shorten the distance between opposite sexes more efficiently than creeping without mucus-trail-following.

The factors that affect decision making about MTF should be investigated in the future, and not only the trail following rate. For instance, the trail following rate of Nodilittorina unifasciata increases when emersed during low tide (Chapman, 1998). Starvation could also affect MTF behavior because some gastropod species ingest mucus as an energy source (Davies & Beckwith, 1999; Hutchinson et al., 2007; Ng et al., 2013). Conversely, animals avoiding the trail of other animals may acquire more food than those that follow others.

MTF reduces distances between members of the opposite sex

Although some studies have suggested that MTF increases the efficiency of mate searching (Ng et al., 2013), no study has produced a definitive demonstration. Previous investigations have focused on the rate of MTF rather than the function of this behavior. In this study, we used an IBM to show that the mean distance between opposite sex individuals was significantly reduced by MTF. However, the distance between males and mutant females was greater than that between males and wild females in 4,907 out of 10,000 simulations. Additionally, distances can be reduced when mutant females follow the mucus of wild males: in such cases the fertilization rate (i.e., reproductive success) of wild males increase even though they do not themselves exhibit MTF behavior.

Locomotion patterns might have affected the distances between males and females in ways other than MTF in our model. The random number generated from the probability distribution for the locomotion distance in H. discus hannai equated to lévy flight (Fig. S3), which is an effective searching behavior when the target is sparsely distributed (Sims et al., 2012). Observed locomotion pattern in another Haliotis species was found to involve lévy flight in the wild (Strain, Johnson & Thomson, 2013); therefore, the natural locomotion pattern of H. discus hannai might also involve lévy flight. The current study focused on MTF as a method of reducing distances between members of the opposite sex; however, other behaviors such as locomotion patterns should also be investigated in the future.

Effect of increased fertilization rate on evolution of MTF behavior

The D allele increased within the population in the MTF model in just over half of the simulations (53/100 simulations). This result indicates mate searching (resulting in increased fertilization rates) could be the evolutionary driving force behind MTF in our model. However, the D allele disappeared from the population in 47 of the 100 simulations, possibly suggesting that the resulting increase in fertilization rate as parameterized is not an evolutionarily stable driving force in MTF. In general, changes in allele frequency depends on the product of the average fitness of the phenotype and the frequency of the allele within the population (Kawata, 1989). When the fertilization rate of the mutant is lower than or almost the same as that of the wild individual, the D allele disappears or does not increase at the beginning of simulations when the allele frequency is low. As described above, the observed distance between males and mutant females was greater than that of wild females in almost half of the simulations (4,907/10,000 simulations): this might explain why the D allele did not always increase early in the simulation. The wild type d allele was not perfectly removed from the population in all simulations (Fig. 2), possibly because the fitness of followed wild individuals was increased by the MTF behavior of mutant individuals without incurring any costs. The d allele might also persist in heterozygous Dd individuals, which express the same degree of MTF behavior as homozygous DD individuals.

As allele D did not increase in about half of the simulations, other benefits in addition to increased fertilization rates could be the evolutionary driving force of MTF in the real world. For example, MTF reduces the energy cost for adhesion and locomotion; gastropods require energy to secrete a mucus trail to adhere and move across the substrate (Davies & Hawkins, 1998). An individual following a mucus trail can decrease the amount of mucus being secreted (Davies & Blackwell, 2007; Hutchinson et al., 2007), thus saving energy. Haliotis discus hannai exhibits MTF in the non-reproductive season, suggesting a possible energetic saving. Unlike the mate-search function, energy savings would accrue continuously.

In the future, the following hypotheses should be tested as factors driving the evolution of MTF as a mate searching strategy. First is that the functional persistence period of trail mucus affects the encounter rate by others. The model’s persistence period of mucus trail was set following observational studies showing that mucus is degraded by bacteria within a day (Herndl & Peduzzi, 1989; Peduzzi & Herndl, 1991). However, the trail mucus in H. asinina could potentially remain on the substrate and be functional for as long as a week (cited as an unpublished observation in Kuanpradit et al. (2012). If the mucus has such a long functional period, abalones could encounter the trails of others much more frequently, which could enhance the distance-reducing effect between opposite sexes. The second hypothesis is the effect of abalone density on the efficiency of MTF behavior. It is possible that abalones are less likely to find the mucus trail of other individuals at low population density, and vice versa. Although the density of abalones was set 1.0 individuals m⁻² in the model, the density varies by year in wild populations (Ohmura et al., 2015). Therefore, the efficiency of MTF behavior could vary by year in the real world, and this should be a focus for future work. In this study, expression of MTF behavior is determined by allele D and d, with nonoverlapping generations; however, the actual relationship between MTF and allele is not clear, and H. discus hannai undergoes repeated spawning for several years. An improved model to address these problems is needed to understand the evolution of MTF behavior.