Keywords

Introduction

Social recognition has rarely received attention in the studies on annelids. This is not surprising since the biology of behavioral interactions in annelids is mostly unexplored. Additionally, there may be a cultural prejudice going back to Darwin that annelids have limited cognitive abilities. Indeed, as Velando et al. (2008) pointed out, Darwin (1871) evaluated the potential for sexual selection to occur in this way: “All these worm-like animals apparently stand too low in the scale for the individuals of either sex to exert any choice in selecting a partner, or for the individuals of the same sex to struggle together in rivalry” (Darwin 1871).

Darwin was convinced of the limited cognitive power of the “the lower classes” of animals, which had “too imperfect senses and much too low mental powers to appreciate each other’s beauty or other attractions, or to feel rivalry.” To Darwin, sexual selection could not work in worm-like organisms, including annelids. In the following pages we advance the opposite hypothesis: since there is evidence that sexual selection is at work in annelids (e.g., Sella 2006; Lorenzi and Sella 2008; Velando et al. 2008), here we suggest that sexual selection may have been a main selective force favoring the evolution of cognitive abilities including social recognition.

The Mechanisms Underlying Social Recognition in Annelids

The physiology underlying social recognition has been under scrutiny in vertebrates in the last decades (e.g., Salva et al. 2012), but is more rarely studied in invertebrates (but see, for example, Tibbetts and Dale 2007). Especially in annelids, our knowledge of the mechanisms at the basis of recognition is very poor. However, many annelids are equipped with an impressive array of sensory structures with which they gain information about their environment. Because the analysis of these structures would go beyond the scope of this book, we refer to the specialized literature for focused overviews (e.g., Purschke 2005). In general, for many marine organisms including annelids, chemical cues and chemoreception mediate many crucial behaviors including defense, reproduction, recruitment and feeding (reviewed by Zimmer and Butman 2000).

In polychaetes nuchal organs are primary chemoreceptive organs, but other complex chemosensory structures have also been described, such as dorsal and metameric ciliated organs and parapodial sensory structures (Lindsay 2009). There is also evidence of specialized visual sensory receptors. Species of the genera Alciopa, Torrea and Vanadis possess image-resolving eyes (Wald and Rayport 1977). Odontosyllis worms have four eyes, which are located two per part on the two sides of the head. The eyes are on lobes that have some degree of movement, they reside in a cavity and have a lens, photoreceptor cells and pigment granules (Wolken and Florida 1984).

Little is known on the proximate mechanisms involved in social recognition in annelids. Linsday (2009) reviewed the main aspects of chemoreception in polychaetes. According to this author, only few pheromones have been identified, which are responsible for mate-recognition and gamete release. In the semelparous species Nereis succinea and Platynereis dumerilii, 5-methyl-3-heptanone induces the nuptial dance. The pheromone triggers its biological effect at different concentrations in the two species, thus ensuring reproductive isolation based on response threshold levels. Different threshold concentrations of the same compound convey different signals in N. succinea as well. Females release cysteine-glutathione-disulfide as a mate-recognition signal which attract males from long distances. A higher concentration of the same compound induces males to release sperm. Other identified pheromones are the egg-release pheromone inosine in N. succinea and the sperm-release pheromone uric acid in P. dumerilii.

Most of these substances are found in many marine invertebrates and their specific biological activity is mediated by concentrations. In this regard an interesting aspect of the sexual pheromones in nereid polychaetes is their activity across species (Hardege 1999; Watson et al. 2003). According to these Authors, extracts from the body fluids of some nereid species (as well as those from other marine invertebrates) induce spawning in other nereid species. Spawning by one species can activate mass spawning events involving multiple species, similar to those observed in the Great Barrier Reef, while coordination of spawning between conspecific partners is controlled by species-specific concentrations of the same compounds (Hardege 1999).

To Which Extent Are These Sensory Mechanisms Involved in Social Recognition?

To our knowledge, most social interactions in annelids occur during mating and reproductive interactions. Worms are able to locate potential partners and recognize each other from worms of other species, because strong prezygotic reproductive barriers impede interspecific mating in annelids. In some species, individuals are not only attracted to their partners, but they also discriminate between mature and non-mature partners or evaluate other, even more subtle, qualities of their partners. Annelids are present in a great variety of ecosystems: there are earthworms, sea worms, abyss worms and fresh water worms. This makes it possible the diversification of the sensory structures according to the environment and the life style of the animals.

Mate Choice Is a Component of Social Recognition

Social recognition is a sophisticated cognitive ability, which is advantageous in many social interactions, from encounters between partners during mating to complex social behaviors (Tibbetts and Dale 2007). Mate choice is a component of social recognition: it involves discriminating between classes of organisms (like conspecifics and non-conspecifics), between sexes (in separate sex species), and between potential partners according to their attractiveness, quality and/or relatedness (Sherman et al. 1997). In this respect mate recognition involves different kinds of class-level social recognition, i.e., recognition of classes of individuals (Gherardi et al. 2012).

Many annelids use chemical and visual cues to locate partners. For example, ripe females of the Odontosyllis species emit light flashes. Flashes might have multiple functions, but some observations suggest that females of the Bermudian fireworms O. enopla attract males by releasing luminescent secretions (Fischer and Fischer 1996).

Males of the marine polychaete worms Ophryotrocha labronica and O. puerilis (the latter is a sequential hermaphrodite) are attracted by water where females have spent some time, suggesting that females might release attractive substances in the water (Berglund 1990, 1991). Some annelids use cues to classify their partners according to mating status, body size, or oocyte ripeness. In dense populations of the hermaphroditic Eisenia andrei redworms multiple mating is common (Monroy et al. 2003). When redworms copulate with mated partners, their sperm compete with those of the previous partner(s) in egg fertilization. During courtship, redworms recognize their partner’s mating status (i.e., whether they have already mated or not and whether they have/have not already received sperm from other partners). Indeed, redworms adjust the amount of sperm they donate to their partner according to the mating status of their partner (Velando et al. 2008). If partners have already mated and have sperm in their spermatheca, redworms transfer three times more ejaculate to their partners than that they do to non-mated partners, thus increasing the chances to fertilize eggs (Velando et al. 2008). This is not the only measure they take to estimate the quality of their partners. They also adjust the amount of donated sperm to the body size of their partner (prudent mating effort, Wedell et al. 2002; Anthes 2010). Large redworms usually have higher fecundity and receive twice the amount of ejaculate than small redworms (Velando et al. 2008).

Eisenia fetida and Lumbricus terrestris earthworms seem to be able to classify partners by body size as they mate preferentially with matching partners (size-assortative mating): variance in body weight and size within a pair is smaller than that between pairs (Michiels et al. 2001; Monroy et al. 2005). For example, L. terrestris use tactile cues during the pre-copulation phase to choose sexual partners. The courtship behavior consists of repeated mutual burrow visits between neighbors and during these visits partners maintain close contact while moving back and forth between their burrow openings. Michiels et al. (2001) advanced the hypothesis that body size is one of the qualities that earthworms assess during these visits to potential partners. Size is used in mate choice and influences the outcome of mating. Earthworms mate sooner with same-sized neighbors than with differently sized ones, and small earthworms visit large neighbors more often than small ones. When mates separate, one of them can be pulled out of its burrow and this mating outcome is more likely to occur to small individuals. Therefore, in L. terrestris, mate assessment is reached by assessment of body size by contact.

Overall, these findings suggest that in earthworms mating is not random and mate choice or ejaculate size are adjusted to partner characteristics or quality. Although the cues that earthworms use to assess their partner conditions (body size or mating status) are largely unknown, these data document that earthworms have an efficient recognition system that can be used for mate evaluation (Domínguez and Velando 2013). Additionally, because there are no processes other than mating which involve interactions between individuals in redworms, these data suggest that sperm competition (a post-copulatory component of sexual selection) might have favored earthworms that have the ability to recognize their partner’s quality, to process and to use this information appropriately. Thus, sperm competition favors earthworms that have some cognitive abilities.

The gonochoric nereidid polychaete Neanthes acuminata provides evidence for sophisticated mate choice. N. acuminata females mate only once, are semelparous and die soon after laying eggs, whereas males care for the developing eggs. As expected in semelparous organisms, females assess male quality very accurately. In particular, since egg survival largely depends on the male ability to care for eggs, paternal egg-caring is a crucial male trait. Fletcher et al. (2009) showed that females classify males based on their experience in egg-caring, and prefer to mate with experienced rather than naive males, even when naive males had low success in male-male fights that often occur in these populations. This suggests that females exhibit mate choice on the basis of the good-parent model of sexual selection (Storey et al. 2013). The choice of the partner is based on transient chemical signals that females recognize as a ‘scent of experience’ (Storey et al. 2013).

In the polychaete worms of the genus Ophryotrocha assessment of mate qualities occurs by means of tactile and/or chemical cues. The sequential hermaphroditic worms O. puerilis are sex-changers: they start their sexual life as males and turn into females when they reach a certain body length (Berglund 1986). Laboratory experiments have shown that females prefer small males as mates and reject large males, possibly to avoid prospective sexual conflicts (Berglund 1990). Large males may be more likely to switch to the female sex and if that happens, the pair will be composed of two females. According to Berglund (1990), size is the crucial cue behind female choice; larger individuals could be recognized as females and smaller ones as males. We do not know how these worms assess their partner body size (e.g., through tactile and/or chemical cues). However, partners may control (or manipulate) reciprocal sex changes (Grothe and Pfannenstiel 1986). When the male partner turns to the female sex, one of the two females now forming the pair will switch back to the male sex. This form of alternating sex change requiring continuous recognition of the current sexual phenotype of the partner is present in other hermaphroditic annelids (Helobdella and Syllis).

The ability to discriminate between the sexual phenotypes of the potential partners (i.e., to perform not an individual- but a class-level social recognition) is present also in the non-selfing, iteroparous, simultaneously hermaphroditic species O. diadema and O. gracilis (Fig. 1.1). In these species, the hermaphroditic phase is preceded by an adolescent male phase. Mature hermaphrodites identify the physiological status of potential partners, and generally form pairs with simultaneous hermaphrodites after a time-consuming courtship, whilst adolescent males are discarded as partners (Sella 1990). In isolated pairs, hermaphrodites regularly take turns in playing either the male or the female role during successive mating events, so that they trade eggs for sperm, and individuals retaliate when partners do not lay eggs in their turn, by stopping egg laying (Sella 1985). Reciprocal egg-exchanges occur only if the same two individuals interact repeatedly and, in theory, such conditional reciprocity requires that partners recognize each other either individually or as familiar partners (Axelrod and Hamilton 1981). Indeed, conditional reciprocity is an evolutionary solution to conflicts over sexual roles in simultaneous hermaphrodites (Leonard 2005), but cheaters exist that repeatedly mate in the male role (Di Bona et al. 2010). In these conditions, selection should have favored reciprocating individuals that identify cheaters and retaliate against them, as was shown by Sella (1988). Reciprocating pairs usually spend most of their time in caring for their egg cocoons, which they often lay in a sort of nest site, where mucous trails are dense. However, occasionally one or both worms leave their nesting site, e.g., for foraging. When nesting sites are unattended, or when only one worm is there, cheaters could enter the nesting sites, pretend to mate in the male role and leave. Cheater success would be limited, and cheater strategy would not spread in evolutionary times, only if partners recognize each other through individual or familiar recognition. In stable pairs, familiarized partners have shorter inter-spawning time intervals than unfamiliar worms have, which suggest some form of partner recognition. Each worm could either recognize its partner through individual characteristics or, more likely, could learn some individual trait of familiarized partners (e.g., odor). There has been only one experimental test for such ability, but the experiment failed to prove that O. diadema worms recognize the individual with which they have previously traded eggs (Lorenzi and Sella 2000). After a period of familiarization between paired partners, partners were separated for a short period, and then pairs were formed again between either familiar or unfamiliar partners. Both kinds of pairs had similar inter-spawning intervals, suggesting that there was no recognition of familiar partners (Lorenzi and Sella 2000). Of course, it is possible that the experimental manipulation was stressful in itself, and worms had to recover from the experimental manipulation before being able to spawn again, irrespective of any partner recognition ability.

Fig. 1.1
figure 1

A pair of Ophryotrocha worms with cocoons containing eggs or larvae

Full size image

Recognition of Social Group Size

Current evolutionary theory makes specific predictions about the abilities of organisms to measure the presence and number of potential competitors over mating. For example, variations in the level of competition for egg fertilization (i.e., sperm competition) should affect the quality and/or amount of resources that males devote to sperm or other male traits (Parker 1998), but this requires that males be able to assess the number of competitors over the male role, i.e., that they recognize their potential rivals, such as mature males, among their conspecifics. Sex allocation theory makes similar predictions: hermaphrodites are expected to adjust their sex allocation to mating group size, i.e., to the number of potential mates (Charnov 1982). According to theory, when the number of mates is small, hermaphrodites allocate few resources to sperm production and devote their remaining resources to egg production. As the number of mates increases, hermaphrodites increase the amount of resources devoted to the male function at the expenses of those available to the female function (Charnov 1982; Schärer 2009 for an updated review). In other words, they adjust their male function to current level of mate competition, trading off resources between the two sexual functions. In experimental practice, mating group size is technically difficult to measure, and therefore social group size is often used as a proxy, under the hypothesis that the larger the social group, the stronger the competition for mating. We are not interested in sex allocation theory here, but in the ability of hermaphrodites to measure the size of their group in order to adjust their sex allocation to social group size.

To adjust sex allocation to current social conditions, hermaphrodites have to perceive the intensity of mate competition they are likely to encounter, i.e., they have to estimate their mating opportunities and change the amount of resources allocated to their sexual functions appropriately. Mating opportunities are likely to increase with increasing group size. Therefore, hermaphrodites might have been selected to be able to measure their group size, i.e., the numerosity of the conspecifics in a population or neighborhood. We recall that in simultaneous hermaphrodites each individual is both a potential mate and a potential rival. Although the ability to assess group size may be common to hermaphrodites in other taxa, in annelids there are a few interesting examples. In the leech Helobdella papillornata, juveniles adjust the volume of their testisac (a sac around the testis) to the social group size they experience during development; the testisac is larger when group size is larger. Here, juveniles use the size of the social group as a proxy for the intensity of sperm competition they will encounter in their adult life (Tan et al. 2004). Hermaphroditic polychaetes of the genus Ophryotrocha can make even more precise estimates of their social group size. Most of these worms live at low densities, but populations vary in size (Simonini et al. 2009). In larger or denser groups, multiple paternity of single egg-clutches is common, as a consequence of multiple matings and competition among sperm of different males (Lorenzi et al. 2014). As adults, these worms adjust their female allocation to social group size (which, again, is likely to be a proxy for mating opportunities). These hermaphrodites estimate how many of them there are, and make this measurement irrespectively of the density of worms (i.e., irrespectively of the number of worms per units of water volume) or of metabolite accumulation (Lorenzi et al. 2005). Experimental evidence suggests that in a few days they can both make these measurements and adjust their female allocation appropriately, but they are also ready to re-adjust it quickly, should group size change again (Lorenzi et al. 2008, Lorenzi pers. obs.). Using female-allocation adjustment as a proxy for the ability of these worms to assess group size, we found evidence that these worms make a more precise estimation of group size for small than for large sizes, but precision is also likely to depend on species-specific conditions under which the different species have evolved. For example, both O. adherens and O. diadema worms allocate their resources to the female function depending on whether they are in groups of 2, 4, or 12 hermaphrodites (Fig. 1.2a). However, the sizes of their adjustments are different. O. diadema hermaphrodites respond more to changes in social group size from 2 to 4 than to changes from 4 to 12 hermaphrodites. In contrast, O. adherens hermaphrodites respond more to group-size changes from 4 to 12 than to changes from 2 to 4 hermaphrodites (Fig. 1.2b). This suggests that worms make more precise assessments of group size within certain ranges of variation than others, either because they are more precise in their estimates of group size or because they are more precise in adjusting their sex allocation appropriately, or both. The perception of social group size is obtained by means of species-specific waterborne chemical cues (Schleicherová et al. 2006). Schleicherová et al. (2010) simulated group-size variations by varying the concentration of the chemical cues that function for group-size assessment. The authors highlighted that O. diadema worms responded to as little as 1 % concentration of the chemical cues produced by 400 hermaphrodites living in 60 ml of water, and noticed that these worms did not make precise sex-allocation adjustments at higher concentration ranges.

Fig. 1.2
figure 2

Perception of social group size and variations in the allocation to female function in focal worms in two polychaete species, Ophryotrocha diadema and Ophryotrocha adherens. a O. diadema worms responded to differences in group size as small as 2 (group size 2 vs. 4) by altering their female allocation in a highly significant way (GZLM, Poisson distribution, identity link function, Wald χ2 = 302.676, P < 0.0001). They also perceived group size differences between 4 and 12, although allocation adjustments were smaller (Wald χ2 = 118.345, P < 0.0001). b O. adherens worms had similar responses, but their female allocation adjustments when group size changed from 2 to 4 worms were smaller than when group size changed from 4 to 12 hermaphrodites (2 vs. 4 hermaphrodites: Wald χ2 = 20.025, P < 0.0001; 4 vs. 12 hermaphrodites; Wald χ2 = 158.455, P < 0.0001). Data from (Schleicherová et al. 2014)

Full size image

The ability to estimate the numerosity of conspecifics has also been shown in the polychaete worm Dinophilus gyrociliatus. This worm has separate sexes, with a marked sexual dimorphism and a peculiar mating system. Females are about 1 mm long, whereas males are 20 times smaller (dwarf males) with no digestive system and a shorter life-span than females. Eggs are laid in rigid egg-capsules which typically contain large eggs (80 μm) which are destined to produce females and small eggs (40 μm) which are destined to produce males (Charnov 1987) (Fig. 1.3). Males and females mate within the egg-capsules, which means that females mate with their brothers. Within each egg-capsule, sex ratio is female biased, because the large eggs destined to produce females always outnumber the small eggs which produce males. However, as local mate competition theory (LMC; Hamilton 1967) predicts, mothers are able to adjust their offspring sex-ratio within egg-capsules to the level of mate competition of the patch where they live. As the number of mothers increases in a patch, mothers increase the proportion of sons among their offspring (Minetti et al. 2013). Like hermaphroditic polychaete worms, these worms also estimate the numerosity of worms that are around in the patch but do not use worm density as a cue (Minetti et al. 2013). The adjustment of offspring sex-ratio by the females of D. gyrociliatus requires the estimate of the numerosity of worms in the patch.

Fig. 1.3
figure 3

An egg-capsule of Dinophilus gyrociliatus with three large female eggs and a small male egg

Full size image

Recognizing Relatedness

Annelids can estimate some level of similarity to their conspecifics, such as the relatedness they share with their partners or whether they belong to the same or a different population.

Earthworms have low dispersal ability and some degree of inbreeding may occur in their populations. Experimental evidence suggests that non-selfing hermaphroditic earthworms E. andrei adjust breeding effort to mate relatedness. Indeed, when they were forced to mate with siblings, they produced smaller egg cocoons than when they were paired to unrelated partners. In this way they reduced their investment in eggs when the likelihood of producing viable offspring (due to inbreeding depression) was low (Velando et al. 2006).

In the polychaete worm N. acuminata, there is evidence that worms changed their behavior when they were exposed to waterborne chemicals produced by conspecifics from other populations, thus demonstrating a role of chemoreception in the discrimination between “home” and allopatric populations (Sutton et al. 2005). This result indicates that N. acuminata is able to perform a sort of chemically based “kin” recognition. Similarly, O. labronica worms discriminate between conspecifics depending on their population of origin, and preferentially court partners from their own population (Åkesson 1972; Lanfranco and Rolando 1981).

Selective Pressures for Social Recognition in Annelids

Although the number of focused studies is low, there are a few that show that some form of social recognition has emerged in annelids (Table 1.1). In some cases, social recognition may have emerged by natural selection. However, in others, sophisticated cognitive abilities like the assessment of social group size (i.e., numerosity) have emerged because they confer mating advantages. Therefore, they are likely to have evolved via sexual selection. We have been made aware from the time of Darwin that sexual selection has been involved as the form of selection favoring a wide range of traits in males, but we now know that it is also at work in hermaphrodites (Lorenzi and Sella 2008; Anthes 2010; Leonard 2006). However, whether sexual selection has promoted the evolution of cognitive abilities has been rarely investigated. In this short chapter, we have shown that sexual selection is likely to be responsible for the evolution of social group size assessment in hermaphroditic  and gonochoric polychaetes, as they respond to group size by varying the proportion of resources they allocate to the male and female function appropriately.

Table 1.1 Evidence for different kinds of class-level social recognition in Annelids and the type of selection which may have driven them
Full size table

Many animals can discriminate the number of conspecifics in their population or neighbourhood. For example, lionesses adjust their agonistic behaviour to the numerosity of their own group as well as of the opposing group in territorial contests (McComb et al. 1994). Guppies and other fish reduce the risk of predation by aggregating in large groups of conspecifics (shoal). For each fish, predation risk is diluted in larger shoals, and evidence has shown that these fish can discriminate conspecifics’ numerosity and prefer joining the larger shoals (Agrillo et al. 2008). In these two examples, the main selective force favouring the ability to discriminate conspecifics’ numerosity is natural selection. Natural selection favours lionesses which make precise estimate of their group numerosity relative to that of the opponent group (thus making appropriate decisions on the costs and benefits of aggressive behaviours). Natural selection (through predation pressure) favours guppies which estimate conspecifics’ numerosity and join larger shoals. In contrast, in polychaetes, the assessment of numerosity is likely to have evolved by sexual selection, because there are no traits that can be crucially affected by the ability to estimate numerosity other than allocation of resources to male and female function.

Recently, Hollis and Kawecki (2014) have documented that cognitive abilities declined in males of Drosophila which were forced to reproduce for more than 100 generations under monogamy. Their cognitive abilities declined more than those of males of polygamous populations that faced both multiple males (i.e., competitors for mating) and multiple females (i.e., mate choice). Potentially, enforced monogamy in simultaneous hermaphrodites could produce similar results after generations, with hermaphrodites reducing their cognitive abilities and losing their ability to estimate group size more than hermaphrodites reared in promiscuity persistently.