Group selection and altruism in animals: possible evolutionary roots

· Ecology and evolution

This work is licensed under a Creative Commons Attribution-Share Alike 3.0 License




Group selection (selection that favours the group over the individual) and altruism (selfless behaviour that favours others sometimes to a degree that is detrimental to one’s own success) have been much discussed in evolutionary theory. Does group selection occur, and is genuine altruism in animals really possible? I shall argue that aggregation (formation of groups) among animals is a very ancient evolutionary phenomenon and widespread among animals. In many cases, selection clearly enhances aggregation over what can be expected on the basis of underlying abiotic and biotic unevenness. Most animal species are rare and live in largely vacant niche space. Crucial for the survival of sexually reproducing species are therefore mechanisms that ensure mating (“mating hypothesis of niche restriction“), i.e., selection for group formation may be vital.Such selection may involvepleiotropic genes (genes involved in the formation of several characters) responsible not only for group formation but also for altruistic behaviour which is essential for the survival of groups.

Aggregation (clustering) in plants and animals

Organisms are distributed in a random, regular (uniform) or aggregated (clustered) way (Figure 1). Random distributions are quite rare but have, for example, been demonstrated for certain parasites of fish. Sometimes, individuals are uniformly spaced; an example is certain bushes in deserts or trees in dense forests, probably as the result of competition for very limited resources ([1]). Most species, however, are clustered (aggregated), as the result of underlying uneven distributions of abiotic resources (such as minerals for plants, or water for plants and animals), or of biotic resources (such as plants for animals, or prey animals for predators), or as the result of grouping in social animals (herds of buffaloes, flocks of birds, schools of fish). In the latter, some selective mechanisms are likely that secure and enhance aggregation. Such mechanisms are, however, not restricted to socially organized species of higher vertebrates, they are likely to occur in many groups of lower invertebrates, particularly in those which reproduce sexually, i.e., depend on meeting mating partners.

Figure 1. Patterns of aggregation: random (constructed from a table of random numbers), regular (uniform), and aggregated (clustered).

Rarity in animals

If aggregation is indeed important for sexual reproduction, we would expect that mechanisms to find mating partners are particularly well developed in those animals which occur at low population densities, that is, are rare. Many studies have shown that rare animal species are much more frequent than common species. For example, according to Krebs (1985) [2], the total number of butterflies caught in a light trap at Rothamstead, England, was 6814 (197 species). 37 species were represented by only a single specimen, 1799 specimens belonged to one very common species, 6 common species comprised 50% of the catch (Figure 2). This example is typical, plots of species against numbers of individuals per species show so-called log-normal distributions: very large numbers of rare and very few common species.

Figure 2. Number of butterflies caught in a light trap at Rothamstead, England. The most abundant species (on the right of the diagram) are not illustrated. Total of 6814 individuals, 37 species only with a single specimen, 1799 specimens belonged to one very common species, 6 common species comprised 50% of the catch. Modified after [2].

Example Monogenea

We use the Monogenea as an example to demonstrate the wide-spread occurrence of aggregation and mechanisms that enhance aggregation. Monogenea are among the best studied Platyhelminthes (flatworms). Most species infect the body surface and gills of freshwater and marine fishes, some are found in amphibians and reptiles, and very few infect marine invertebrates or have adopted an endoparasitic way of life (see Monogenea).

Aggregation and rarity

Aggregation in host species and host individuals

Concerning parasites in general, aggregated distributions are the rule: usually few host individuals are heavily infected with parasites, whereas most hosts are either not infected at all or carry a very light parasite load. This applies to Monogenea. Not considering host individuals but host species, we find that few host species contain the largest number of parasites, whereas most host species contain few (Figure 3). In other words, as in other animal groups, rarity of species is the rule and not the exception.

Figure 3. Abundance (mean number of parasites of all species per host species) of metazoan ectoparasites on heads and gills per species of marine teleost ( 5666 fish of 112 species). Modified from [3].

Host and site specificity

A monogenean species is almost always restricted to a single or a few related host species, and usually microhabitats (organs or tissues or parts of them) on the host are restricted to various degrees. A typical example is illustrated in Figure 4. The only monogenean found on 15Oligopletes saliensin Brazil occurred only on the basal parts of the gill filaments. Current competition cannot be the cause of this restriction, because no other parasites (except for 2 copepod individuals) were found on the other parts of the gills or the body surface. Therefore, there must be a genetic basis for the restriction. Selection must have led to niche restriction in the course of evolution.

Figure 4. Ectoparasites on the body and gillls of 15Oligopletes saliensin Brazil. P= pseudobranch, 1-4 gills 1 to 4. Short arrows: a species of Monogenea, Large dot (long arrow) copepodOrbitacolaxsp., small dot (long arrow) copepodBliassp.. Modified from [4].

Proximate causes of specificity

Kearn in particular, has investigated various direct (proximate) causes of host specificity in monogeneans (review in [5]). In some species that have been investigated, hatching of larvae is induced by specific factors in the host’s mucus, in others, an endogenous hatching rhythm corresponding to the activity patterns of the host guarantees that the correct host is infected. In species of the very speciose genusGyrodactylusdifferential mortality after infection determines host specificity; worms attach themselves readily to incorrect hosts but die on them. Other species of the same genus can survive but not reproduce on the wrong hosts. Immune factors may be involved. Thus, certain species live longer in mucus from susceptible than from immune hosts. In some tapeworms, suckers used for attachment to the intestinal villi fit exactly unto the villi of certain fish species only. In the wrong fish attachment is impossible. A similar mechanism may apply to some monogeneans, although it has not yet been demonstrated.Sexuality in MonogeneaMorphological evidence on its own strongly suggests that sex is of extreme importance for the Monogenea and that sexual selection has been an important factor in their evolution (Figure 5). Thus, male and female copulatory organs of many species are of considerable complexity and apparently operate in a key and lock mechanism (although few observations of copulation have been made).

Figure 5. Male and female copulatory organs of Lamellodiscus squamosus(A) andL. major(B) from the gills ofAcanthopagrus australisin SE Australia. Modified after[6].

All Monogenea are hermaphroditic, but self-fertilization is prevented (or at least made difficult) by morphological (and other?) characteristics. For example, the male and female copulatory organs of many monogeneans are extremely complex and located in such a way that self-fertilization seems almost impossible (Figure 6). Furthermore, studies of other parasitic flatworms (the aspidogastreanLobatostoma manteri) have shown that, although introduction of the male cirrus into the uterine opening (through which sperm is introduced) of the same worm has been observed, in single-worm infections egg cells were haploid and died at a few-cell stage. In other words, sexual reproduction involving two mates is important and probably essential.

Biological function of specificity

Above discussion suggests “ultimate” (biological) causes of specificity (proximate causes are direct physiological or other causes of a pattern, ultimate causes are causes related to the biological function of a pattern that have arisen through evolution). We have seen above that most monogenean species are rare, that is, live in largely vacant niche space, and that they depend critically on sexual reproduction involving mating partners. Selection to ensure mating is therefore of primary importance. Rohde has established the “mating hypothesis of niche restriction” which claims that an important factor responsible for niche restriction is the enhancement of opportunities to secure cross-fertilization. The hypothesis is supported by the observation that niches of asexual stages are often wider than those of sexually mature stages; that niches become narrower at the time of mating; and that microhabitats of sessile species and of species with small population sizes often are narrower than those of non-sessile species and of species with large population sizes.Monogenean species inhabiting different parts of the gills often differ markedly in the structure of their adhesive organ (the haptor or opisthaptor). Clamps on the haptor may be of different sizes and shapes (Figure 6) and this may explain why certain species can only attach themselves to particular portions of the gills. Since restriction to small microhabitats enhances mating chances, sexual selection may have led to differentiation of the adhesive disk.

Figure 6.Heteromicrocotyloides mirabilis. Note the adhesive disk carrying five kinds of clamps, some flattened and functioning as suckers (on the left of the disk), others consisting of two valves that grasp gill filaments. The male copulatory apparatus (cirrus and accessory copulatory organ) is of great complexity and located ventrally. The vagina is located dorsally. Modified after Fig.2 in [7].

Alternative explanation: competition

Interspecific competition could, in principle, explain why species of Monogenea have particular microhabitats into which they have been “pushed” by competitors. However, as shown above, we find extreme niche restriction even in parasites that have no competitors whatsoever (Figure 4). Of course, the possibility must be considered that competitors may have existed in evolutionary history but have become extinct for one reason or the other. However, no evidence points to this possibility and it is not likely, because cases of niche restriction in the absence of competing species are very common. – Concerning intraspecific competiton for resources, this certainly will occur, but only when population densities of mongeneans are high. As we have seen, however, population densities of most monogenean species are very low, and resources (blood, epithelial cells, on which worms feed) are in unlimited supply as long as the host is alive. Therefore, neither inter- nor intraspecific competition are likely to play an important role in niche restriction.

Implications for group selection and altruism

Group selection assumes that selection for the good of the group (the species or population) and not of the individual (or gene) can explain certain evolutionary patterns. For example, it has been suggested that animals somehow “agree voluntarily” to scale down reproduction in order to overcome overcrowding (by so called epideictic displays of Wynne-Edwards [8]), or that animals reduce intraspecific aggression avoiding death of the contrahent, because it is in the interest of the species (Konrad Lorenz [9]). Dawkins [10] has given valid arguments against the possibility of group selection: if a single individual opts out of this “strategy”, i.e., puts his own interests first, his genes would soon predominate and wipe out the others.According to Ruse [11], from a biological viewpoint altruism means “doing something to help the reproductive chances of someone else, even though apparently this entails the diminution of one’s own chances. It is obviously the sine qua nonof social behaviour…”. Group selection could easily account for such a mechanism, but individual selection, as assumed for example by sociobiology, cannot so easily do this. Sociobiology [12] has, however, suggested three important possible causes of altruism among animals: kin selection (it is in one’s interest that relatives reproduce in order to guarantee continuation of one’s genes), parental manipulation (parents manipulate an offspring in such a way that he/she helps their other offspring at the expense of its own reproductive success, very similar in some respects to kin selection), and reciprocal altruism (one helps others because one hopes that others will help you, as in cleaning symbiosis). All these mechanisms can be explained by individual selection.We ask whether the prevalence of group formation in the animal and plant kingdoms generally and monogeneans specifically can in some way be explained by selection for group formation (SGF) and perhaps form the basis for altruism. After all, if groups are essential for sexual reproduction involving two partners, mechanisms responsible for group formation may include mechanisms which are altruistic. At least in higher organisms, groups would fall apart if some degree of altruism was not involved. As stated by Ruse, altruism is ” obviously thesine qua nonof social behaviour.” I do not claim altruistic behaviour for the monogeneans, but simply suggest that group cohesion has evolved early in evolution and, because it is so prevalent, may form the basis of altruism in higher animals like humans.Does a genetic mechanism exist that links altruism so closely with group formation that selection promotes altruism “without knowing it”? It selects pleiotroic genes for group formation that also determine altruistic behaviour, although the latter is detrimental to the individual carrying the genes; group formation is more important to the individual than the risk it may run by becoming altruistic.Of course, any gene and not only those involved in group formation could conceivably be pleiotropic having effects on altruism, however, such effects would be irrelevant in non-social species and therefore not express themselves. – If true, we would have to add a fourth mechanism to the three listed by Ruse: group formation. It should be understood, however, that this is not the same as group selection.Cooperative breeding and eusociality, evidence for group augmentation?, although not restricted to birds, has been particularly well studied in birds by many workers, using theoretical and experimental approaches. In cooperative breeding, not only the parents but related and sometimes unrelated individuals of the same species help in looking after the nest, eggs and offspring, i.e. appear to show altruistic behaviour. Some studies have shown that environmental factors, such as shortage of territories or unpredictability of resources, predispose birds to cooperative breeding, although environment is only one of the factors involved.

Numerous studies (e.g., [13], [14], [15]) deal with the genetic basis of cooperative breeding. Kinship (inclusive fitness) may often give a satisfactory explanation, but is less satisfactory when non-related individuals act as “helpers”, and there is increasing evidence that non-related helpers are indeed quite common. Therefore, explanations based on group augmentation (increase in the density and size of the population) are attempted, because it seems obvious that the size of a group enhances its genetic diversity and thereby its evolutionary survivability [16]. However, group augmentation is difficult to disentangle from other explanations. In the words of Wright (2007 [17]), “group augmentation thus becomes a ‘rubbish bin’ – a last-resort theory in cooperative breeding, which is invoked (but not properly tested) when other …….. hypotheses such as kin selection or pay to stay fail to explain our results”.

In eusociality (formation of ‘superorganisms’, ‘states’), as found for example in bees, ants and termites, the ‘altruistic’ behaviour is far more advanced than in cooperative breeding. Most of the individuals of an insect ‘state’ have lost their ability to reproduce altogether and devote their entire energy to supporting the colony. Until recently, the commonly accepted explanation was based on kin selection theory, i.e. the concept of ‘inclusive fitness’ (an ant worker is more closely related to the queen than it would be to its offspring, hence it is in its interest to secure the production by the queen of as much offspring as possible). Recently, however, it has been proposed that kin selection theory, although sometimes applicable, is not necessary to explain eusociality [18]. According to this new model, several steps, all explained by classical natural selection, are necessary on the way to eusociality: 1) groups within a population are formed; 2) genes are accumulated that enhance the chances to become eusocial (cooperation in food gathering, tunnelling and guarding of resources, nest building by a female and protection of larvae by her); 3) genes arise that prevent dispersal from the nest of the female and her offspring. – Important in the context of our discussion, according to this model, group formation is indeed the essential first step in the origin of eusociality.

Altruism in humans

This discussion has dealt exclusively with non-human animals. Humans differ from other animals in possessing a huge and complex brain that has permitted development of culture. Ethical convictions usually involving altruistic principles are part of culture and transmitted by tradition. Although genes certainly provide the general basis for culture formation in being responsible for development of the brain in the first place, it is unlikely that they are solely (if at all) responsible for each cultural trait. Dawkins, for example, claims that genes are not significant in the formation of human cultural traits, some sociobiologists at least attribute a greater role to them. Is it possible that only humans can truly act in an altruistic way, for example because they are able to arrive at ethical philosophical principles as in Buddhism, Hinduism or Schopenhauer’s philosophy? Peter Mersch has attempted to give a generalized interpretation of evolution including that of eusociality (state formation in animals) and human culture using a systems approach (also here, here), which provides new and refreshing insights into how ethical principles may have evolved. In particular, he provides valid arguments against Social Darwinism (the strongest is always right).


Robert MacArthur and Joseph Connell (1966). The biology of populations. John Wiley &Sons, N.Y., London, Sydney.Charles J.Krebs (1985). Ecology. 3rd ed. Harper & Row Publ. N.Y.Klaus Rohde (1998). Latitudinal gradients in species diversity. Area matters, but how much? Oikos 82, 184-190.Klaus Rohde (1991). Intra- and interspecific interactions in low density populations in resource-rich habitats. Oikos 60, 91-104.Klaus Rohde (1993). Ecology of Marine Parasites 2nd ed. CAB Internatinal, Wallingford, OxoF.R. Roubal, F.R. (1981). The taxonomy and site specificity of the metazoan ectoparasites on the black bream, Acanthopagrus australis (Gunther), in northern New South Wales. Australian Journal of Zoology, Supplement 84, 1-100.Klaus Rohde, (1977) Habitat partitioning in Monogenea of marine fishes. Zeitschrift für Parasitenkunde 53, `171-182.V.C. Wynne-Edwards (1962). Animal Dispersion in Relation to Social Behaviour. Oliver and Boyd, Edinburgh.Konrad Lorenz (1966). On Aggression. Methuen, London.Richard Dawkins (1976). The Selfish Gene. Oxford University Press, Oxford.Michael Ruse (1979). Sociobiology: Sense or Nonsense? D. Reidel Publishing Co. Dordrecht, Boston, London.Edward O. Wilson (1975). Sociobiology. The New Synthesis. The Belknap Press of Harvard University Press. Cambridge Mass., London.Andrew Cockburn (1998). Evolution of helping behavior in cooperatively breeding birds. Annual Review of Ecology and Systematics 29, 141-177.Ralph Bergmüller, Andrew F. Russell, Rufus A. Johnstone, Redouan Bshary (2007). Integrating cooperative breeding into theoretical concepts of cooperation. Behavioural Processes 76, 61-72.Ralph Bergmüller, Andrew F. Russell, Rufus A. Johnstone, Redouan Bshary (2007). On the further integration of cooperative breeding and cooperation theory. Behavioural Processes 76, 170-181.Hanna Kokko, Rufus A. Johnstone and T.H. Clutton-Brock (2000). The evolution of cooperative breeding through group augmentation. Proceedings of the Royal Society B 268,187-196.Jonathan Wright (2007). Cooperation theory meets cooperative breeding: exposing some ugly truths about social prestige, reciprocity and group augmentation. Behavioural Processes 76, 142-148.Martin A. Nowak, Corina E. Tarnita and Edward O. Wilson (2010). The evolution of eusociality. Nature 466, 1057-1062. ———– Copyright Klaus Rohde

%d bloggers like this: