With a Special Section on Primates

Sexual Dimorphism: "Occurrence of populations where individuals differ in respect of two distinct sets of phenotypic characters, sex itself being one of them." (The Penguin Dictionary of Biology 1992)

Size: "The relative bigness or extent of a thing, dimensions, magnitude." (The Concise Oxford Dictionary 1990)

ABSTRACT

In most animal species, the males and the females are of different sizes. This Sexual Size Dimorphism (SSD) has been hypothesized to have evolved for a number of different reasons. The initial dimorphism of egg and sperm concerning size, nutrient content and motility, subjects the two sexes to different selection forces. The female is commonly the larger sex, but in mammals and birds the male is most often larger. Females are believed to be larger in most animals because egg production increases with size, while males are believed to be larger in mammals and birds because large size gives an advantage in acquiring mates. Other hypotheses are also examined, including ecological niche differentiation and allometric relationships. Finally, recent literature dealing with the evolution of sexual size dimorphism in primates is reviewed.

INDEX

WHAT IS SEXUAL SIZE DIMORPHISM (SSD)?
- Measurement of size
- Calculating SSD
GENETICS OF SSD
THE EVOLUTION OF SEX
THE ONTOGENY OF SSD
NATURAL SELECTION
- Fecundity selection
- Parental care
- Ecological niche differentiation
- Defense against predators
- Starvation
- Thermal stress
- Loading constraints
- Dominance over the other sex
- Differential flight requirements
SEXUAL SELECTION
- Male competition and female choice
- Mating System and SSD
- Protandry
- Sperm competition
ALLOMETRY
CONFLICTING SELECTION FORCES
SELECTION FORCES WORKING IN CONCERT
SEXUAL SIZE DIMORPHISM IN PRIMATES
CONCLUSIONS
REFERENCES

WHAT IS SEXUAL SIZE DIMORPHISM (SSD)?
Males produce sperm, females produce eggs. But besides than this definitional difference, there are other characters that constitute 'femaleness' and 'maleness' of animals. Darwin (1871) divided sexual dimorphism into three subgroups: (1) Primary sexual characters - differences in the reproductive apparatus, i.e. gonads, their ducts and associated glands. (2) Secondary sexual characters - differences in all other physical characters as well as in behavior. (3) Ecological sex differences - differences in niche, brought about by physical or behavioral differences. Sexual Size Dimorphism, SSD, is a difference in a physical character other than the reproductive apparatus, namely the difference between the sexes in absolute size. In most extant species the female is probably the larger sex, but in mammals and birds the case is commonly the reverse (Andersson, 1994). SSD can be found in many different scales. The extreme on one side are the dwarf males in some endoparasitic barnacles which are so small that the females first were mistaken to be hermaphrodites. On the other hand, in elephant seal the male is more than five times the weight of the female (figure 1). This difference in size between the sexes, and its causes and effects, has provoked a lot of research.

Figure 1. Female and male elephant seals, Mirounga angustirostris. (From Trivers, 1985)

Measurement of size
Size can be measured in many different ways, and depending on what one measures, different SSDs can be calculated (e.g. Oxnard, 1983). Body mass can be measured by using weight or body volume. Weight and volume of individuals vary, however, due to various things such as nutritional status, season, etc. so at different sampling occasions, data on the same individual can differ. Measurements of several individuals over a longer period in time are therefore sometimes undertaken, calculating mean values for individuals or populations based on a number of samples. Even then, size can vary between the populations of the same species because of size-specific mortality, population-specific growth patters, and the availability of food.

Other more static measurements of endo- or exosceletal animals can be gathered by measuring skeletal parts which do not vary over time with nutritional status, such as length or thickness of a bone, a shell, or total length of the organism. For example, commonly used measurements in birds are tarsus length and beek depth because of their correlation with general size and their non-fluctuating character. Such data are then taken as a sample of the absolute size itself and used as such, or the absolute size is calculated using a formula.

Some organisms continue to grow after maturity and some even continue to grow all their lives. In such cases it is important to know where in the life cycle the measurement should be taken as size in such cases is age-specific. Some cases of dimorphism reported in the literature has been shown to be due to sex differences in age in the sampled animals, and not because of any general sex difference in body mass (Stamps, 1993; Stamps et al. 1994). One study by Howard (1981) on bullfrogs, for example, revealed a significant SSD in the population. When checking for age, however, this dimorphism didn't exist, both sexes having the same adult growth rate. The discrepancy was due to a higher mortality for larger males. Thus, from a physical point of view, the dimorphism did not exist, even if it did exist from a population point of view. A similar situation appears when comparing breeding animals where the maturation age for the sexes are different. This then also becomes a comparison between animals of two different mean age groups (Howard, 1981; Halliday and Verrell, 1986). Care also has to be taken to see that the sample one takes is typical for the whole population. For example, models made for Anolis lizards show that males often are larger than predicted by female size distributions within the same sample. This is hypothesized to be because at high quality areas where most individuals are found, small size males have difficulty of holding a territory (Stamps et al. 1994). To prevent such errors knowledge of growth trajectories can make it possible to calculate the probable size distribution in a sample and then see if sampled values fit predicted values. Also, age specific measurements can be made at maturation, around the time of mating, at final size (if there is any), or of the growth trajectories themselves (Stamps, 1993; Stamps et al. 1994).

Calculating SSD
The classic way of calculating SSD is by using the ratio between the size value of one sex divided by the other. As pointed out by Ranta et al. (1994), this can cause some statistical irregularities if used with the wrong statistical method, since ratios of normally distributed variables, such as size, are usually leptokurtotic and skewed to the right. Departures from normality can also occur under some conditions which, however, can be corrected with arcsine or square root transformation, or avoided by using a non-parametric method. Ranta et al. recommend the use of regression residuals as an option. Using ratios can also obscure allometric relationships between the sexes, as pointed out by Fairbairn and Preziosi (1994). SSD commonly increases with size in animal species with larger males than females, while the trend the reverse in species with larger females than males. Although this might seem contradictory, it is part of the same trend, as discussed below. In some species sexual selection seems to be selecting for large male size, even though the male is the smaller sex (e.g. Andersen, 1994; Fairbairn and Preziosi, 1994). When plotting weight of one sex against ratio, however, the resulting curve is non-symmetric around zero, while plotting the weight of the sexes against each other makes the relationship clear (see Fig XX below). Logarithmic values may also cause statistical irregularities. For example, when one predicts values using a least squares regression analysis on logarithmic values, and then deconverts them, these predicted values will be underestimated (Smith, 1993).

Lovich and Gibbons (1992), set up four criteria which a SSD variable should meet; (1) it should be properly scaled, (2) it should have high intuitive value, (3) it should produce values with one sign, (positive) when sex A is larger than sex B, and the opposite sign when sex B is larger, and (4) it should produce values that are symmetric around a central value, preferably zero. The criterions excludes ratios and logarithms, two often used calculations. Their own suggestion is (size of largest sex divided by size of smallest sex) +1 if males are larger or -1 if females are larger, "arbitrarily defined as positive when females are larger and negative in the converse situation" which in my opinion falls on their own criterion of intuitiveness and also, because of the arbitrariness of the calculation, it isn't mathematically consequent. The best suggestion is probably to adjust the calculation to the question and statistical method to be used.

GENETICS OF SSD
The genetic basis of metrical characters, including size, is usually polygenic (Lande, 1980 and references cited therein). Selection on one such character in one sex usually produces a correlated response in the other sex. For example, Eisen and Hanrahan (1972), selected for sexual dimorphism in dimorphic mice (males>females) in two different manners. Two lines were selected for large females and small males, two lines for large males and small females. As expected, in the first case the dimorphism decreased and in the second the dimorphism increased, although the results were not straightforward. Interesting was that when selecting for larger females, the males also increased in size, while selecting for smaller females gave smaller males. This to the degree that the males in the line where larger females were selected for ended up larger than in the line where large males were selected for. This was explained by maternal effects, and shows how cautious one has to be when dealing with size selection. High genetic correlation between the sexes can greatly slow the rate of evolution of sexual dimorphism (Lande, 1980).

If there is a directional selection for large size on a polygenic character in a species, how come variation isn't depleted? One answer was found by McLain (1987), who determined the heritability of size in Nezara viridula, southern green stink bug. It was shown that there was a significant heritability of size, but also that size variation in nature probably was due more to environmental conditions, such as food availability. The size of an individual was thus determined more by non-heritable factors, and could therefore not be transmitted to the next generation. In such a case selection may not be consequent enough to deprive the population of its variation. Another way in which directional selection can even work for larger variation is if there are different reaction norms to food availability in the population. For example, one genotype may result in larger individuals than another in a food rich environment, while the relationship is the opposite in a food scarce environment. In this case, in a fluctuating environment selection for large size will favor different genotypes at different times. On the other hand, in stable habitats, which genotype is favored will depend directly on food availability, and populations may then become fixed in different states in different locations. Providing there is some migration between populations, variation will thus not be depleted in these.

It is probable that sexual dimorphism in higher animals is most often not dependent on genes present in one sex and not the other. Rather, it is the expression of certain genes which is sex-limited. This expression of genes depends on sex-specific hormones produced by the gonads, which in turn are determined by sex-linked genes. (Andersson, 1994). Rice (1984), came to a different conclusion and, in a theoretical analysis of the effect of X-linkage on the evolution of sexual dimorphism, stated that "sex chromosomes facilitate the evolution of sexual dimorphism and that X-linked genes have a predominant role in coding for sexually dimorphic traits".

In a population genetic model Lande (1980) found that when under no sexual selection, the joint evolution of the mean phenotypes of the sexes increases the mean fitness of a population. When under sexual selection, however, each sex achieves a local optimum phenotype. Sexual selection can this way produce a large deviation of the mean phenotype from the species optimum and thus cause a substantial fitness decrease in the population. Maynard Smith and Brown (1986, cited in Andersson, 1994) hypothesized that there is a selection for larger males in many mammal species, increasing the species mean size through genetic correlations, and this way increasing the risk for extinction. The fossil record suggests that most mammal lineages do increase in size over time. In fact, the increase of size in phylogenetic lineages is so common that it has been named "Cope's law" of phyletic size increase (Pilbeam and Gould, 1974). Since mammals on average do not increase with time, this implies that extinction is more common in large species (Andersson, 1994).

THE EVOLUTION OF SEX
Sex is believed to be about 1500 million years old (Stearns, 1987). The reasons it evolved and is maintained is a debated subject which will not be discussed here, however, its appearance is a prerequisite for sexual dimorphism. Sex can be subdivided into three components: recombination, reproduction and gender. Sex is not necessary either for recombination or reproduction; these occur in asexual organisms as well as sexual organisms. Gender, however, used here as a sort of 'maleness' or 'femaleness' of an organism, is a major and unseparable, however not necessary, consequence of sex. Stearns (1987) believes the historical sequence for the evolution of sex to have been: (1) asexual reproduction; (2) limited recombination (bacterial); (3) meiosis; (4) mating types; (5) anisogamy; (6) gender and; (7) incompatibility types. It is in steps five and six where SSD comes into the picture, the sexes before this being morphologically similar. Anisogamy is the production of gametes of unequal sizes, and this is where the concept of males and females first occurs. By definition, females is the sex which produces large gametes, and males is the sex which produces small gametes. The evolution of anisogamy has occurred many separate times; isogamy is found today only in the certain protoctists and fungi (Margulis and Sagan, 1984, 1985). Anisogamy evolves because the production of gametes of different sizes give different advantages: small gametes can be produced with little investment and therefore in abundance while large gametes provide nutrition for the zygote which then has a better chance of survival (Stearns, 1987; Maynard Smith, 1978). Small gametes can have increased motility by investing in more powerful tails and smaller size while large gametes can sacrifice their motility and increase their longevity and nutritional supply by increasing in size (Halliday, 1994). A zygote formed by the fusion of the two gamete types is both more likely to occur as well as survive than a zygote formed by the fusion of two gametes of the same type and anisogamy would therefore be a stable state (Parker et al., 1972). Selection will also favor that only gametes of the two different types fuse with each other (Halliday, 1994). What follows is also that no gamete of a third size can invade the population and that the fitness of the two sexes is equal only when the sex ratio is 1:1 (Maynard Smith, 1978).

All other sexual dimorphism is based on the initial dimorphism of eggs and sperm. As soon as anisogamy once has evolved, different selective forces act on females and males. Then the evolution of gender differences can begin, i.e. the differentiation of the sexes regarding other aspects than the gametes. The difference in size and motility between the two types of gametes provides for the fact that sperm is most often delivered to the egg, even if there are exceptions such as seahorses and pipefish (Miller and Harley, 1992).

One consequence of the difference in energetic cost of the gametes, and thus greater investment of females in their offspring, is that the males will have energetic potential to compete more for access to females (Trivers, 1972). Whereas the number of offspring a female can have is related to the number of egg which can be produced, the fitness of the male depends on how many females he can fertilize. Thus the males can be expected to compete more for the females. This does not necessarily mean that the male will become the larger sex, as large female size can be advantageous for other reasons. According to Fisher's (1930, cited in Trivers, 1972) theory for sexually reproducing organisms, an equal parental investment on the male and the female progeny will be favored by natural selection. In species with SSD this leads to the dilemma of how to invest enough in the large offspring and at the same time to avoid investing too heavily in rearing offspring in that particular sex (Trivers, 1972).

THE ONTOGENY OF SSD
SSD does of course have both ultimate and proximate causes. Understanding of SSD requires knowledge of both the fitness of growth patterns and the finite size of adults, as well as why there is a difference in the first place. More commonly the cause is of genetic origin, and therefore ultimate in nature, but males can, for example, invest more in displays, contests or searches for females, the differential energy allocation causing the males to be smaller than the females, in spite of there being no genetical difference (Ghiselin, 1974, cited in Andersson, 1994).

Kozlowski (1989), described three potential routes during ontogeny for one sex to become larger than the other. (1) The larger sex may develop from larger eggs. In this case there is a difference in cost between the two sexes, which could result in the production of a larger number of the relatively cheaper offspring. (2) The larger sex can grow for a longer time period. This has implications on the probability of survival to maturity since larger animals commonly require longer time to mature. The reproductive life-span of the larger sex will because of this also be shorter than that of the smaller sex. (3) The larger sex can grow faster. Since there is a difference in the first place, this third causation model implies that there are some disadvantageous effects of faster growth on mortality and/or life span. In birds however, a study by Teather and Weatherhead (1994), indicated that it was the smaller sex in the species that grew faster relative to final size, and thus reached its final size quicker. No adaptive explanations was necessary to explain the differing growth rates. Teather and Weatherhead concluded that both sexes grew at maximum rates determined by their eventual body size. Shine (1990), in an analysis of data from reptiles, amphibians and fishes, found that the sex which matured at a larger size in 90% of the cases also was the sex which attained the larger average size. This suggests that dimorphism is determined mainly by differences in growth rate and maturation age. Continuous growth affects the degree of the dimorphism, but probably doesn't change its direction (Anderson, 1994).

Howard (1981), found a significant SSD in a population of bull-frogs. This dimorphism didn't exist when checking for age, both sexes having the same adult growth rate. The discrepancy was due to a higher mortality in larger males, who consequently matured earlier as to have a longer reproductive life-span. Apparently the sexes had the same advantage of large size, males for defending their territories, females for fecundity reasons. Knowledge of age and growth trajectories cannot be neglected if one wants to measure SSD (Stamps, 1993; Stamps et al. 1994).

NATURAL SELECTION
"Most of the distinctive characteristics of higher taxa are surely adaptations that evolved by natural selection..." (Futuyma 1979: page 434, cited in Coddington, 1990)

"All of biology compels us to recognize that organisms are not optimally designed, that many features are not adaptive, and that species may differ for reasons other than natural selection."(Futuyma 1986: page 254, cited in Coddington, 1990)

Fecundity selection
In the majority of animal species, except mammals and birds, the female is larger than the male (Andersson, 1994). This is believed to be because egg production increases with body size, but also because the keeping of eggs requires more space then the keeping of sperm. While the females need to use much energy for the eggs and therefore may need to build a larger energy storage onto themselves, the males can use the energy more for moving around, or emerging early as a mean of inseminating more females. In mammals and birds the male is usually the larger sex, but there is still a high positive correlation between maternal body mass of homeotherms and mean mass of both the individual progeny and the entire litter (Cabana et al. and references cited therein, 1982). There exists numerous studies in many taxa which show that female fecundity increases with size (e.g. Howard, 1988b; Harvey, 1990; Ralls, 1976; Wiklund and Karlsson, 1988). Shine (1988), however, points out that whereas clutch size often is related to female size in many animals, this doesn't mean that the life-time reproductive success for larger females is larger than that of smaller females. This is simply because the growth to larger size takes time and energy which could instead have been invested in offspring. A life-history perspective is needed to answer the question of the relationship between size and fitness properly.

Fecundity selection may also act on female size in the opposite direction. Small females may be able to devote more of their energy intake into reproduction (Wiley, 1974, cited in Andersson, 1994). Small female size may also be favored by earlier breeding possibilities per se, e.g. r-selection (Stearns, 1992).

Parental care
In animals with developed parental care, the fecundity may not depend so much on egg numbers as on the ability to nurse the offspring to independence (Trivers, 1972). Ralls (1976), suggested that the parental care improves with size in mammals. The ability to win competitions over food or other resources may be associated with this (Trivers, 1972). Jönsson and Alerstam (1990), however, found in a study of shorebirds that in species where the female was smaller, she had the main responsibility for parental care, while in species where the male was smaller the roles were reversed. They suggested that small size was more adaptive for the sex mainly responsible for parental care since this sex then needed less energy to maintain itself. In some seahorses and pipefish the male nurses the eggs. In spite of this male brooding, in one of the examined pipefish species (Nerophis ophidion), which was dimorphic, male fecundity was not correlated to size. Female fecundity, however, still was. In the other species (Syngnatus typhle) both males and females had increased fecundity with increased size, and maybe as a consequence this species lacked size dimorphism (Berglund et al., 1986).

Ecological niche differentiation
Size has many implications on the way animals meet the environment and each other (Peters, 1983). Thus, in species exhibiting SSD, males and females would be expected to interact with the environment differently. There are numerous studies showing ecological differences between males and females (Shine, 1989 and references cited therein). The difficulty is to determine if such differences are the cause of the SSD or just a result of it. Ecological causation models imply that to lower the intraspecific competition, selection has acted on the difference between the sexes directly, or different specializations have evolved due to different ecological needs. Ecological causes should not influence the direction of sexual dimorphism, but once one sex differs from the other they can influence the degree (Shine, 1989, 1991; Madsen and Shine, 1994). This makes these type of models harder to test. The only reliable evidence of the validity of the hypothesis has been suggested by Selander (1972), to be dimorphism in trophic structures larger than that expected by body size differences, and in a direction inconsistent with sexual selection. These criterions would rule out sexual selection as cause of the dimorphism and at the same time give support to the idea that the trophic structures are dimorphic because of divergence of diet, and not only as a by-product of size itself. On the other hand, these criterions are so rigid, they probably fail to identify a lot of cases.

In a review of evidence supporting the ecological causation model, Shine (1989), finds that dimorphism in trophic morphology exists in all animal phyla. In the majority of these cases it is impossible to conclusively exclude other causes. He found dwarf non-feeding males without functioning mouth parts to have evolved in many taxa, as well as males and females with differing mouth parts because of different diets. This dimorphism has then probably not evolved because of competition over resources between the sexes, but as independent specializations due to differing nutrient requirements between males and females. Shine (1991), later applied Selanders (1972) criterion of dimorphism in trophic structures larger than expected by body size, to snakes. The dimorphism in head size gives no reproductive advantages in snakes, and is therefore to be expected to be caused by dietary divergence. Shine infers that the sexes originally didn't diverge because of ecological factors, but when the divergence in body size once had happened, independent adaptations for foraging evolved. One study by Forsman (1991), showed a geographic variation of size in adders which was due to a significant correlation with prey size. There was also a correlation between body size of the male and the degree of sexual dimorphism. He could not, however, find any indications on ecological niche differentiation, but instead argued that the larger size of the females could be due to fecundity selection. In another study, on Arafura filesnakes, Houston and Shine (1993) did, however, find different foraging specializations of the sexes. Males, which were smaller, ate many small prey at a time, while females generally ingested one large prey. The specialization was still there when snakes of the same size were compared. This is consistent with the predictions made for the ecological causation hypothesis, but fecundity selection has not been excluded as an explanation.

Slatkin (1984), analyses the theoretical part of the ecological causation hypothesis in a quantitative genetic paper. He divides the ecological causes into three classes of assumptions. (1) Different selection pressures. If males and females have different social roles or different energetic needs to ensure successful reproduction, the sexes can have different optimal weights. (2) Two or more size optima for the species. Assuming a species has two size optima, one sex could evolve to one optimum, and the other to the other. (3) Competition between the sexes for access to limited resources. This would lead to the sexes diverging in size, making it possible for the sexes to exploit different recourses. All three of these hypotheses are shown to be theoretically possible causes for sexual dimorphism.

Ecological competition is predicted to be more pronounced in monogamous, territorial species where sexual selection is weak and the pair probably would be each others worst competitors. This contradicts, however, the trend of more dimorphism in more polygamous species (Searcy, 1979). That is not to say the ecological competition model is proven wrong, just that this piece of evidence points to sexual selection as a more common cause of dimorphism.

Defense against predators
If one sex has the main responsibility to protect against predators, it would benefit by larger size. Anderson (1986), for example, found that SSD is larger in primate species or populations with higher predation rates. Weaponry such as horns or molars would also be predicted to be selected in this model. Since an allometric connection between the size of body parts and total body size most often exists, selection for large weapons could also act as a selection for large absolute size. Weaponry is, however, more often associated with intraspecies competition than with predator defense (Andersson, 1994). Thus, selection for large size through weaponry would be more properly categorized as sexual selection.

Starvation
Large size can be advantageous during food shortages as storage capacity increases faster with body size than does metabolism (Downhower, 1976). Per unit body mass, larger animals have smaller requirements than do small and can thus survive a longer period of time on internal energy stores than smaller animals (Peters, 1983). Millar and Hickling (1990) therefore argued that larger animals will be selected for in unpredictable environments because they can endure longer periods of starvation due to larger energy reserves, and that small size will be favored only by a constant lack of food because of the lower cost of maintenance. Millar and Hickling's conclusion was disputed by Speakman (1992) who argued that in an unpredictable environment there is selection on size both during periods of food abundance and during starvation. Thus no one direction of size selection can be predicted in unpredictable environments. One could, however, picture a "bottle-neck" scenario, where animals who could not survive periods of starvation would be selected against while the selection during high food availability would not be quite as severe (Sillén-Tullberg, pers. comm.). Speakman also argues against the general rule that small size would be favored in a predictable low-food environment, he points at the need to be large to defend resources and that large animals have larger alimentary tracts and hence capacity to process food. It has been argued for some Antarctic birds and northern fur seals that the need to hold territories for a long time without eating requires large body size for the males, who consequently stock up large reserves (Downhower, 1976).

But if large size is advantageous when withstanding starvation, why then do the larger males in dimorphic species usually have higher mortality than females during food shortages (Berteaux, 1993, and references cited therein)? One suggestion is that because sexual selection has selected for larger males, females are closer to the ecological optimum for the species (Gaulin and Sailer, 1985). If this is the case, the more ecologically stressed sex can probably withstand less pressure in the form of starvation. Males, at least in the juvenile phase, have been hypothesized to have higher nutrient requirement (Clutton-Brock et al., 1985). Weatherhead and Clark (1994) argue, based on their data on red-winged blackbirds, that higher mortality in males not necessarily is contrary to the selection for large size on them. The mortality in the study was higher in general for males because when competing for territories, males had to encounter harsher weather than the females. But because larger males had higher probability of survival, there was no counter-selective force on large size. The data suggested that the limit on male size instead lay in the sub-adult stage during migration.

Thermal stress
Greenwood and Wheeler (1985), have proposed a theory they call the "hot blooded hypothesis" to explain why males are larger than females in mammals and birds. They cite several studies to show that females are more likely to experience heat stress than males, and that this stress causes damaging effects on fetuses. Therefore the females are proposed to be under selective pressure to avoid damaging temperature elevations during the reproductive period. In mammals males have external testes and in birds the testes are located close to the posterior air sacs into which air is first drawn, while the reproductive tract of the female is internal with lesser cooling possibilities. The testes can thus be effectively cooled, so males would not be subjected to temperature selection. There would, however, be a thermoregulatory advantage of being of smaller size for females since the temperature conductance is related to surface area to volume ratio. Smaller females would thus have more effective cooling. The authors argue that although birds have mostly external embryo development when behavioral adjustments will thermoregulate the eggs, the early stages takes place internally. The hot blooded theory is consistent with the fact that size dimorphism with larger males is a general case only in endotherm animals.

Loading constraints
Which sex that carries the other during mating can also have an effect on dimorphism, as well as being effected by it (Fairbairn, 1990 and references cited therein). Fairbairn's study on temperate water striders supported causation model, the larger sex carrying the smaller. Greenwood and Adams (1987), in a criticism of a paper by Ward (1986, cited in Greenwood and Adams, 1987) argues that the male biased SSD in Gammarus pulex is due to the male carrying the female during copula, and that even if male-male competition is important it doesn't explain why the males need to be larger than the females. Ward (1987) defends his conclusion by showing that size is important in G. pulex male-male competition, and stressing that all size combinations of male-female copulation had been observed and that this implied that small males had no problem in carrying the large females. Wiklund and Forsberg (1991), found no support for the loading constraints hypothesis in a comparative study of 23 species of butterflies.

Dominance over the other sex
For different reasons, one sex may dominate the other through its larger size. It has, for example, been suggested in birds of prey that it is beneficial for the female to be larger than the male to protect the young from cannibalism, to maintain pair bonding, or to dominate the male to provide food (Mueller and Meyer, 1985). For species where males commonly force copulations, such as semiaquatic bottom-walking turtles, there is also an advantage for the male of being relatively larger than the female (Berry and Shine, 1980).

Differential flight requirements
Studies on insects suggests that flight velocity, flight duration and migration distance increase with body mass. Since such flight characteristics often is important for both sexes in insects, Fairbairn (1990), suggested that this would tend to reduce dimorphism. If the sexes have differing flight requirements, however, SSD could evolve because of this. McLachlan (1986), investigated the rain-pool dwelling midge, Chironomus imicola, and found that their flight was dimorphic. In investigating the cause of this, he found that to disperse the eggs, the females often had to fly many kilometers with a full load of eggs. This in contrast to the male flight patterns where the adaptation was more towards short (45 minutes), acrobatic flight during male competition in mating swarms. The males thus had an advantage of small size, increasing their agility, and the females of large size, increasing their stamina. McLachlan from this argued that the protandry in this species, caused by shorter larval development time in males, could be a device for achieving small size, rather than increased available time to mate. In birds of prey, differential flight requirements have also been discussed as a causation model, the males being hypothesized to be selected to be more agile for aerial displays, nest defense and/or hunting (Mueller and Meyer, 1985).

SEXUAL SELECTION
To explain sexual dimorphism, especially when some traits that males have acquired reduce survival and therefore work against natural selection, Darwin 1871 put forward the theory of sexual selection. The idea was, however, already with him in 1859 when he wrote that sexual selection, in contrast to natural selection, 'depends not on a struggle for existence, but on a struggle between the males for possession of the females; the result is not death to the unsuccessful competitor, but few or no offspring' (Darwin, 1859, p.136). Generalized in the finished form to concern both sexes, sexual selection 'depends on the advantage which certain individuals have over other individuals of the same sex and species, in exclusive relation to reproduction' (Darwin, 1871, p.256).

Most commonly sexual selection acts on male size. Andersson (1994), in a review of published data did, however, find 24 cases of sexual selection on females. Almost all of these were due to male choice, males preferring large, fecund females as mates. He also found some role reversed cases with females competing over males, having advantage of large size in the contests. The existence of sexual selection in female animals is a neglected area of research even though such selection can exist under many circumstances (Arnold, 1994). More common, however, are the cases in which sexual selection is believed to work on male size. This is believed to be the typical case for birds and mammals, but it also occurs in other taxa such as for example amphibians, reptiles and crustaceans (Darwin, 1871).

Male competition and female choice
Sexual selection is thought to operate on males characters in two main ways: male competition where the winning male gains access to more matings, and female choice where the female chooses the male she deems more fit. In his review of sexual selection, Andersson (1994), found 48 cases where the relatively larger males had an advantage in dominance contests. For example in snakes, where females commonly are larger than males, the instances where the males are larger have been shown to be correlated with the amount of male combat (Shine, 1978). One study by Anderson and Fedak (1985) on gray seal males went one step further and investigated the connection between body size, dominance and fecundity. Even though the male seals formed no linear dominance hierarchy, large males lost fewer encounters than small males. The largest breeding males were rewarded by up to ten times the progeny compared to the smallest breeding males. There are few studies which clearly show that dominance also lead to higher production of offspring (Andersson, 1994). Ellis (1995), however, reviewed nearly 700 studies, most giving indirect support to the hypothesis that dominance increases reproductive success. For females it is often possible to reliably determine number of offspring, but for males, which is the sex where dominance contests most often are thought to influence fecundity, indirect measurements often are the only data available. Some of the weaker indicators of fecundity for males in the reviewed studies have not been tested to be reliable, and some of the stronger indicators, such as number of copulations, are not straightforward either. For example in macaques, paternity determined with DNA fingerprinting revealed that there was no significant correlation between male copulatory frequency and fecundity (Stern and Smith, 1984, cited in Ellis, 1995). Ellis review does however, give support to the connection between dominance and fecundity, but a large proportion of the investigated species divert from this picture. This is especially true of primates which is also the most thoroughly investigate group.

Females can choose males for a number of fitness-related reasons, such as the ability to hold territories, forage for food or defend the young. In mottled sculpins, for example, there is a fecundity advantage of large males because they can better defend the eggs in the territory against egg-eating fishes of the same species (Downhower et al., 1983). Females should therefore tend to choose large males to mate with. Male mottled sculpins are, however, also able to ingest small females. There should therefore be an avoidance of large males by the females. Also, females should avoid large males because these males can be suspected to have been chosen by a larger number of females, and thus the territory would be unable to hold more egg-masses. For whatever the reason, the females in this species tend to choose males who are a certain size ratio in relation to themselves. This selective choice from the females keeps the selection pressure on the male population towards a constant ratio compared to the female population. Interesting is that although there was a fluctuation in absolute size over different mating-seasons, the difference between male and female size was constant.

Both male competition and female choice can also select for smaller males. For example, Payne (1984) hypothesized that, since lekking bird species with 'reversed' dimorphism all have aerial display, agility rather than large size is favored. An example of female choice of small males are Drosophila subobscura, where small males are better at following the females in courtship dancing. Another female choice criterion in D. subobscura, however, is the size of a nutritious drop which the males offer the females and which increases with size (Steele and Partridge, 1988).

Mating System and SSD
There is assumed to be a general correlation between SSD and the amount of polygyny exhibited in a species. This has been shown to be the case in for example pinnipeds, ungulates and primates (Alexander et al. 1979), large herbivores (Jarman, 1983), and birds (Orians, 1969). The reason for this correlation is that the selection pressures for males to grow larger than females are greater when the possible advantage is greater. This is the case if, for example, the males are polygynous and the access to females is dependent on the size of the male, either by male-male competition or female choice. In many mammals this is the case and then the males have to show off their size and/or fight in order to get to inseminate females. In this case the selection pressure for large size is large on the males and SSD can readily develop. The causal connection between a polygynous mating system and SSD is so accepted that dimorphic fossil species are said to be polygynous with the size difference as indicator. This is not necessarily the case though, since many dimorphic species are monogamous (Rowell & Chism, 1986), but in most species the relationship holds.

Björklund (1991) argued that to be able to show the causal connection between mating system and dimorphism, it is not enough to show the association in existing species since this can be due to phylogenetic inertia ultimately caused by some other factor. A more correct approach is to show that after a shift in mating system, a shift in selection pressure occurs, this way increasing or decreasing sexual dimorphism. Another criterion increasing the credibility of the investigation is to see that the change in SSD comes after the change in mating system. He also added a third criterion: to be able to show that the connection is not due to genetic drift. In a phylogenetic analysis of grackles, Björklund was then able to show sexual selection on tail size, but not on body size. This type of comparative studies is powerful in determining cause and direction of SSD. Commonly, early studies used species as single data points (e.g. Alexander et al., 1979), methods insensitive to phylogenetic inertia and therefore sometimes erroneous. Nylin and Wedell (1994) points out, however, that the conclusions of many of these studies are still valid, and that results obtained with phylogenetic methods also have to interpreted with caution.

An extreme case of skewed mating success for males are the lek-breeding birds. Strong SSD would be predicted since male fecundity is extremely dependent on competition. In most of the lek-breeding bird species dimorphism is the rule, but nearly one quarter of the species are monomorphic in both plumage and size (Payne, 1984). Höglund (1989), did a thorough phylogenetic analysis to see if phylogenetic inertia would explain this, but he found that no conclusion can be drawn from phylogeny as to whether dimorphism has succeeded lek or vice versa. Oakes (1992), reanalyzed Höglunds data and found a significant relationship between lekking and SSD, concluding that lekking promotes SSD and that SSD is higher in lekking taxa. Oakes conclusion was, however, due to an error when using the comparative phylogenetic method. Höglund and Sillén-Tullberg (1994), showed in an analysis of the reanalyzes that Höglund's original conclusion holds. Another hypothesis to explain the occurrence of monomorphic lekkers implies a trade-off between morphological and behavioral responses to sexual selection. The dimorphic species, however, also have elaborate vocal and other displays (Trail, 1990). If all monomorphic species were cryptic, then predation avoidance could be considered, but this is not the case. Trail (1990), instead suggested that monomorphism, which sometimes involves exaggerated plumage, is due to intrasexual competition in both sexes.

Sexual selection can work also in a monogamous mating system, if mates differ in quality or if the sex ratio is skewed (Darwin, 1871). One example was given by Price (1984b), in a study of Darwin's finches. He found that the sex ratio during the study was 3 males to 2 females because of differential mortality during a drought. Although monogamous, the males thus had to compete over females, larger males having larger reproductive success. The males did not compete for females directly, but for size of territories. The larger territories that large males could hold were hypothesized by Price to indicate male quality to choosing females, another possibility being that large territories were good per se.

Protandry
The need for one of the sexes to emerge early in order to have a longer reproductive period and/or to be first on the mating field, has also been hypothesized to cause SSD. Early male emergence has been shown to have an effect on how many females they get to inseminate in some butterflies (Wiklund and Fagerström, 1977). Here the selection is on males to emerge earlier from the pupal stage and they consequently don't have the time to grow as large as the females who have no advantages of earlier emergence. This can also give rise to selection for smaller relative size in males who then can emerge earlier (Singer, 1982). Wiklund and Forsberg (1991), however, found that in 8 directly developing butterflies there was no relationship between larval feeding time, nor time spent in the pupal stage with size dimorphism. In diapausing species there was a difference, hypothesized to be because of the lower phenotypic plasticity of the post-diapause pupal development. In another study of one species of butterfly, Pararge aegeria, which has both protandrous and non-protandrous populations, similar dimorphism was found in both types of populations (Nylin et al., 1993). No relationship between protandry and dimorphism thus existed in this species. Nylin et al. argued that size and development time could be optimized independently, to a limit of course. One earlier study suggested a trade off between large male size and early emergence, large size being favored (Wiklund, et al., 1991).

Sperm competition
Selection acting on a physiological trait, such as weapons, can have implications for body size. A special case of sexual selection is sperm competition which can be connected to body size dimorphism if there is a correlation between large body size and sperm production (Harcourt et al., 1981) or nuptial gift production (Wedell, 1993). There exists an allometric relationship which, however, also is true for many organ weight to body weight ratios. Males can compete with sperm simply by having relatively larger testes than their competitors. Only when testes or nuptial gift size is wholly dependent on body size can therefore sperm competition be assumed to cause SSD.

One case where sperm competition is assumed to play a role for size dimorphism is in fishes, where species with internal fertilization (e.g. poeciliids) often have smaller males than females, while commercial fish with communal spawning (e.g. clupeids) approach monomorphism (Parker, 1992). Some insects also show a strong correlation between male body size and spermatophore size, for example butterflies (Svärd and Wiklund, 1989) and bushcrickets (Wedell, 1993). In primates, sperm competition selects for larger relative testes size, testes being larger in species where females mate with several males than in species where the female mates with one male (Harcourt et al., 1981).

ALLOMETRY
Why does SSD increase with size in animal species with larger males than females? Why is the trend the reverse in species with larger females than males? Although this might seem contradictory, as shown by Fairbairn and Preziosi (1994), it is part of the same trend. Namely, greater evolutionary divergence in male than female size and strong covariation of size between the sexes (figure 2). This is consistent with the fact that in some species sexual selection seems to be selecting for large male size, even though the male is the smaller sex (e.g. Andersen, 1994; Fairbairn and Preziosi, 1994). When selection acts for larger or smaller size in one sex, genetic correlations can lead to a change in size also in the other sex. Larger absolute size may be permitted by lower predation risk, fewer competing large species and higher energetic efficiency in larger animals. Size and weapons may also be more usable in competitions among larger individuals (Andersson, 1994). The correlation of SSD and size may, however, also be non-adaptive.

Figure 2. "Allometry for SSD based on the general allometric model, female size=a(male size)b. The allometric exponent, b, becomes the slope of the regression of log (female size) on log (male size). The dotted line with a slope of 1.0 illustrates the degree of SSD. If females are larger than males and b<1.0, SSD declines as size increases (hypoallometry). If males are larger than females, b>1.0 yields a positive correlation between size and SSD (hyperallometry). Throughout the full range in size of males and females, D(male size)>D(female size). Regression of log (male size) on log (female size) would yield the same conclusion, but b would be greater than 1.0." (From Fairbairn and Preziosi, 1994).

In a phylogenetic study of water striders, Andersen (1994), concluded that there existed a basic female biased dimorphism but also a negative allometry where males grew relatively larger as size increased. Andersen argued that this wasn't necessarily a causal relationship, but could not provide any evidence for sexual selection, nor phylogenetic constraints, instead emphasizing the complexity of the issue. Björklund (1990), in a phylogenetic study of birds, found that the evolutionary change in mating system was connected to a change in absolute size, but that after correcting for this size change, there existed no residual dimorphism difference between monogamous and polygamous species. Even without correcting for size, there was only a weak correlation between mating system and size dimorphism.

CONFLICTING SELECTION FORCES
"The greater the degree to which males depart from the presumably optimal size of the females, the more poorly adapted they should be for general existence, unless by their increased size they are able to exploit food resources not available to the smaller females." (Orians 1969)

Even when large size is advantageous in order to win mates, something sets the upper limit. Different selection forces sometimes act on SSD in concert, other times in different directions. A certain male size may be favoured by female preferences and male-male competition, while a smaller size would be favoured by survival selection. There would then be two selection forces woking in different directions on male size, resulting in an intermediate size distribution (figure 3). In kudus for example, males are larger than females because of reproductive benefits, but mortality for fully grown males is higher than that for females, mainly due to predation. Male kudus, however, have an additional benefit of size during food shortages (Owen-Smith, 1993). In a study of 28 North American passerine birds, male mortality has been shown to increase with the relative size compared to the female. The larger the difference in size between the sexes, the larger the difference in mortality. This was also true for another sexually dimorphic character, namely plumage coloration, where brightly colored males have a larger mortality than the more cryptic females (Promislow et al., 1992). In anurans, although it has been shown that both male-male competition and female choice favors larger males, females are larger than males. There is a clear correlation between female size and fecundity, but this does not explain why the males are small. Woolbright (1983) argued for a model where the costs of advertising, maintaining territories and lower food intake in species with prolonged breeding periods affects the size of males. Sullivan (1984), however, found Woolbright's assumptions suspect and when testing the model on some anuran species he found that the data did "not strongly support [Woolbright's] hypothesis".

Figure 3. Graphic representation of male size distribution as a function of conflicting natural and sexual selection. (Modified from Maynard Smith, 1982, cited in Andersson, 1994)

One thoroughly examined example are the small males of some spiders. It was suggested by Darwin (1871) that the males benefited from being small by avoidance of being eaten by the female. Although the males in such spider species are much smaller than the females, males still often fight over females and the largest male wins, sexual selection acting on larger size of the male (Elgar, 1991). A game theoretical model was presented by Vollrath and Parker (1992) to explain why the size difference still persists. Since males and females have so different life-styles, males ceasing to build orb-webs and instead spending their time locating, courting and mating with females (Elgar et al. 1990), male mortality is higher than female. Thus the sex-ratio becomes female biased, this way making male competition over females less important. The fitness of the male then becomes more connected to early emergence, and the consequential smaller size, as this increases the males chance of survival to mature age. As this happens, more and more males survive to maturation, this way increasing the male to female ratio so that male competition, and thus also relative male size, again becomes more important. This model was supported by field data on the spider Nephila clavipes. One interesting detail is that size dimorphism is significantly less pronounced in species where the mating takes place on a mating thread as compared to when the mating takes place in the hub of the net. This is hypothesized to be an adaptation to avoid the sexual cannibalism by females of males (Elgar, 1991). Madsen and Shine (1994), examine another exception to the 'general' rule that in species with male combat, the male is larger that the female. In adders, females are larger than males despite the fact that males wrestle each other for access to females, the larger male almost always winning. Madsen and Shine's hypothesize that (1) there is a higher between mating mortality in females than in males, because females on average mate every second year, males every. Females also suffer higher mortality in years when they are reproducing than in years when they are not. (2) Thus the number of reproductive events is lower in females than in males, which means that (3) the first year of reproduction makes a larger contribution to life-time reproductive success in females than in males. Consequently it is important for the females to be as large as possible at the first mating, so that fecundity the first year is maximized. Although growth continues, SSD established at maturity generally changes very little (Shine, 1990).

Arak (1988), put forth a mathematical model of selection for dimorphism along the lines that if the function for males and females relating body size to survival is the same, then the difference in body size between the sexes is proportional to the difference of their respective selection gradients. He then proceeded to test the model on anurans and found that it explained a great deal of the variation of SSD. A problem Arak faced was that although size had greater influence on male than female reproductive success, females were still larger. This, he stated, could be due to (1) erroneous data, partly since some measurements were done without checking for age, (2) male and female mortality function related to size were not equal, and (3) the indeterminate growth of anurans which makes it difficult to know at which stage one should measure the dimorphism.

SELECTION FORCES WORKING IN CONCERT
A study on Darwin's finches (Price, 1984a) showed two selection pressures working on SSD in the same direction; the females having an advantage of being small because of earlier breeding possibilities, males having a competitive advantage of large size. Large size may be of importance to both sexes, but the question is sometimes to which sex is it more important? In a work on parasitic wasps, for example, it was shown that larger size relative to the own sex was advantageous to both species. Females of the category 'small' had a third of the reproductive prospects compared to those of category 'large', while in males it was half (van dem Assem et al., 1989). The same direction of selection was thus working with different strengths.

Sexual selection can also work in concert with ecological selection for wider foraging niches for the species. Erlinge (1979) studied this in weasels where the male is larger than the female. Movements and behavior of the males during the mating season suggested a polygynous mating system where there was competition for females. But at the same time, during certain parts of the year the males switched to a seasonally available prey and were also shown to take prey of different size. Thus two selection forces seemed to select for increased SSD; sexual selection and ecological niche differentiation. However, when the sexes were sharing the same habitat, outside the mating season, their diet was the most similar. This suggested that ecological niche differentiation in this case had little or no causal role in the evolution of SSD.

SEXUAL SIZE DIMORPHISM IN PRIMATES
"The following important generalizations about sexual dimorphism [in primates] have emerged from ... quantitative studies: (1) It is always lacking in species living in monogamous family groups; it may or may not be present in species where males have breeding access to more than one female (polygyny); (2) It tends to increase with increasing body size; (3) It is especially prevalent among terrestrial primate species; (4) It is virtually absent among living prosimian primates, uncommon in New World monkeys and markedly present only among Old World monkeys and apes."(Martin 1980)

The ultimate causes of SSD in primates does not differ from that of other mammals, but it is a thoroughly investigated group. There have been several hypotheses put forward to explain why males are larger than females in most species, and why females are larger than males in some. The most often investigated hypotheses are, as summarized by Ford (1994): (1) Natural selection-fecundity related to body size and energy allocation strategies may differ between the sexes. (2) Sexual selection-focuses mainly on male competition over females. (3) Allometry-for some reason or other, SSD increases with body size. (4) Phylogenetic inertia-SSD being retained from an ancestor who was dimorphic. (5) Differences in diet-different diets have different energy content. There are also varying costs in acquiring dietary items. As each sex tries to optimize its size/energy intake/energy use level, this might lead to increased/decreased SSD. (6) Differences in habitat-whether the species is arboreal or terrestrial has been shown to correlate with SSD. (7) Predator defense by males-males are large to discourage predators.

More than 100 years after Darwin's theory of sexual selection a large comparative analysis of 100 primate species was done and a correlation between SSD and socionomic sex ratio (number of males/female) was shown to exist, as sexual selection theory predicts (Clutton-Brock and Harvey, 1977; Clutton-Brock et al., 1977). A correlation was, however, also found between SSD and size indicating allometric causation. As the authors pointed out, socionomic sex-ratio might not be the best variable to measure male-male competition. This is particularly true in multi-male mating systems where a marked dominance hierarchy exists. The study also indicated that SSD was slightly less pronounced in arboreal than terrestrial species and in nocturnal than diurnal species, but revealed no connection to diet.

Leutenegger (1978, 1982), argued that the previous study had not properly explored the connection between polygyny and body size. His analysis indicated that the allometric relationship between size and SSD was valid only for polygamous species, while this relationship did not exist in monogamous species. The model Leutenegger suggested is that polygyny selects for size connected with SSD, while monogamy promotes monomorphism and tends to limit body size. Additional studies by Leutenegger and Cheverud (1982, 1985) gave more support to the hypothesis. Here the authors tested the correlation between SSD and several different possible causing factors. Body size was found responsible for 83% of the variation in SSD, while the next largest factors, mating system was responsible for only 6.8%, and diet 2.5%. Thus, they argued, sexual selection is a negligible force compared to the influence of SSD. These studies excluded species which had larger females than males because of the method chosen of measuring dimorphism: log (male weight - female weight). This calculation was chosen for other statistical reasons, but since the logarithms of zero and negative numbers don't exist, species with reversed SSD had to be excluded.

Gaulin and Sailer (1983), included more species in an analysis and looked a little closer on what happens when one measures dimorphism using the formula log(male weight - female weight). This measure gives an equal degree of dimorphism to a species where the male is 55 kg and the female 50 kg, and to a species where the male is 5.5 kg and the female 0.5 kg. So by using this measurement weight is given too much statistical importance. Gaulin and Sailer instead recommend using log(male weight) - log(female weight) instead, since ecological and sex roles are more likely to be connected to relative size differences. Also pointed out was that Leutenegger and Cheverud regressed weight on dimorphism to get their 83% explanation. Then they regressed the other variables on the residuals of weight dimorphism after the influence of weight had been removed. This gives a bias towards weight which has very little to do with the analyzed data. Gaulin and Sailer then examined the relative contributions of allometry and sexual selection, and showed that size alone adds little explanatory power aside from its contribution via the interaction with mating system. On the contrary, the contribution of mating system was found to be large, thus supporting the theory of sexual selection. Later, Ely and Kurland (1989) also reanalyzed Leutenegger and Cheverud's work and, predictably, found the same problems as Gaulin and Sailer (1983) did in their examination, but added "deficiencies in selection of a structural path model" to the list. Ely and Kurland claimed that their purpose with the study was to call attention to the autocorrelation method itself, which they felt has been neglected in a review of comparative methods by Bell (1989, cited in Ely and Kurland, 1989).

"We may often falsely attribute to correlation of growth, structures which are common to whole groups of species, and which in truth are simply due to inheritance; for an ancient progenitor may have acquired through natural selection some one modification in structure, and, after thousands of generations, some other and independent modification; and these two modifications, having been transmitted to a whole group of descendants with diverse habits, would naturally be thought to be correlated in some necessary manner." (Charles Darwin 1871, p. 185)

To analyze the role of phylogenetic inertia, Harvey and Clutton-Brock (1985), used nested ANOVAs to locate the taxonomic level at which most of the variation is found, and thus identify the level most appropriate for analysis. They found that for many continuous characters, such as body size, 85% of the variation is found at the subfamily level. Many of their analyzed life-history characters were correlated with body size, but they made no analysis of SSD. Cheverud et al. (1985, 1986), however, used a "network autocorrelation approach" to try to sort out the phylogenetic inertia in SSD causation. They found that about 50% of SSD variation was due to phylogeny and 36% was due to size or scaling while an "only negligible proportion of the variance" was due to mating system. Cheverud et al. did not, however, take any regards to ultimate causes of the phylogenetic inertia but submitted "that only the covariance of [species] specific values... may provide evidence of an adaptive evolutionary relationship between species' attributes". Again log(male weight - female weight) was used as measure of dimorphism why the influence of absolute weight is biased. The choice of this measure was motivated by the fact that it shows a linear response to a constant linear selection pressure.

One prediction that can be made if sexual selection is the cause of SSD is that males should have evolved to be larger, not females to be smaller, since it is on males the increased selection pressure lies. The probable route to SSD was examined by Willner and Martin (1985). Brain size is more constant than body size throughout evolution. Since brains in female dimorphic simians are larger than expected while male brains of dimorphic species are the same relative size as those of monomorphic species, Willner and Martin argued that SSD has evolved through body size reduction in the females, thus speaking against sexual selection. This is in line with a study by Pickford (1986), who proposes a selection of small females because of their way to exploit the environment, particularly food resources. The female is proposed to have to reduce her body weight so that her effective body weight (mothers weight plus infants weight plus extra component because of lactation) becomes equal to the males. This hypothesis would also explain why primates who make use of more nutritious food are less dimorphic. In a later study, Martin et al. (1992), again argued for selection acting on smaller female size. This time with the added arguments that in terrestrial ecosystems where predation is higher, there is increased need for rapid reproductive turnover, thus selecting for smaller females who can reproduce faster, this way explaining the correlation between habitat and SSD. Gaulin and Sailer (1985), found the opposite relationship in a non-phylogenetic comparative study. Males deviated more from the species optimum and they therefore argue that SSD has evolved through body size increase in the males. In their analysis, diet quality gave the strongest prediction of species body size, but in species with polygamous mating system the males deviated more from expected weight than in monogamous species.

In a study aimed to examine the general influence of predation on primate evolution, Andersson (1986), in a non-phylogenetic test, found that SSD is larger in species or populations with higher predation rates, which is in accordance with the hypothesis that males are larger than females to protect against predators. Predation is, however, highly dependent on habitat, terrestrial species being more preyed upon than arboreal.

Some studies of SSD have been done on limited groups of primates, and some on single species. Kappeler (1990, 1991), in an extensive study of prosimians, concluded that they generally lack SSD. In the species where SSD does occur, Kappeler could not find any connection to either absolute size or phylogeny. He points out, however, that in simians below 3 kg there also is no connection between size and SSD, so the allometry may be non-existent or not detectable in primates of these sizes. Thus the allometric connection may still be there. Kappeler's data also shows that there is no connection between mating system and dimorphism in prosimians, while the correlation is significant in simians. Instead Kappeler argued for a combination of selective agents to account for patterns of dimorphism in prosimians. These agents include absolute body size, female fecundity and male agility. Ford (1994), did a similar study on platyrrhines where, however, she found that the main causes of the dimorphism was to be found in sexual selection, but also in dietary influences. Body size, phylogeny and habitat could not be connected to SSD. An extreme exception to the connection between SSD and polygamy in primates is De Brazza's Monkey, Cercopithecus neglectus (figure 4). This species is not only monogamous, but would have been extremely dimorphic even if it was polygamous (Leutenegger, 1982). In an attempt to explain this phenomenon, Leutenegger and Lubach (1987) took a closer look at this species, and found three possible explanations: (1) De Brazza's monkey is not strictly monogamous, (2) the SSD is probably an effect of phylogeny and, (3) the tendency towards monogamy is probably a result of the selection of small group size due to anti-predator reasons. According to sexual selection theory, polygamous human cultures should exhibit more SSD than monogamous cultures. That is, if marriage cultures correlate with mating cultures. Gaulin and Boster (1992) investigated 93 societies, half monogamous half polygamous, to see if this was the case. They could, however, find no such correlation.

Figure 4. SSD in primates is mostly correlated with mating system, such as in the mainly monomorphic gibbons (right) who are monogamous. There are, however, exceptions such as De Brazza's monkey Cercopithecus neglectus (left) who, despite the fact that males are substantially larger than the females, are mostly monogamous. (From Harvey et al. 1987).

"About the only commonly accepted observation is that dimorphism in the Primates is widespread and varied." (Pickford 1986)

CONCLUSIONS
"Sexual dimorphism is such a widespread phenomenon throughout the animal kingdom that there are almost certainly a variety of reasons for its occurrence." (Slatkin 1984)

In most animal species, the males and the females are of different sizes. This SSD has been hypothesized to have evolved for a number of different reasons. The initial dimorphism of egg and sperm concerning size, nutrient content and motility, subjects the two sexes to different selection forces. High genetic correlation between the sexes, however, slows the evolution of dimorphism. Commonly the female is the larger sex, but in mammals and birds the male is most often larger. Females are believed to be larger in most animals because egg production increases with size, while males are believed to be larger in mammals and birds because large size gives an advantage in acquiring mates. Other hypotheses, some specific to certain species or lineages, are also hypothesized to cause SSD in certain cases. These include ecological niche differentiation, predator defense, advantages of large size during food shortages, thermal stress, loading constraints, a need for one sex to be dominant over the other, differential flight requirements, protandry, sperm competition and allometry. Different selective forces sometimes work against each other, stabilizing a certain size, sometimes in concert, leading to extreme SSD such as dwarf males.

REFERENCES
Alexander, R. D., Hoogland, J. L., Howard, R. D., Noonan, K. M., Sherman, P. W. 1979. Sexual dimorphisms and breeding systems in pinnipeds, ungulates, primates, and humans. Pages 402-435 in N. Chagnon and W. Irons, eds. Evolutionary biology and human social behaviour: an anthropological perspective. Massachusets: Duzbury, North Scituate.

Andersen, N. M. 1994. The evolution of sexual size dimorphism and mating system in water striders (Hemiptera: Gerridae): a phylogenetic approach. Ecoscience 1: 208-214.

Anderson, C. M. 1986. Predation and primate evolution. Primates 27: 15-39.

Anderson, S. S. and Fedak, M. A. 1985. Grey seal males: energetic and behavioural links between size and sexual success. Animal Behaviour 33: 829-838.

Andersson, M. 1994. Sexual selection. Princeton University Press, New Jersey.

Arak, A. 1988. Sexual dimorphism in body size: a model and a test. Evolution 42: 820-825.

Arnold, S. J. 1994. Bateman's principle and the measurement of sexual selection in plants and animals. American Naturalist 144, Supplement: S126-S149.

Assem, van dem, J. Iersel, van, J. J. A. and Harthog, los-den, R. L. 1989. Is being large more important for female than for male parasitic wasps? Behaviour 108: 160-195.

Berglund, A., Rosenqvist, G. and Svensson, I. 1986. Mate choice, fecundity and sexual dimorphism in two pipefish species (Syngnathidae). Behavioural Ecology and Sociobiology 19: 301-307.

Berry, J. F. and Shine, R. 1980. Sexual size dimorphism and sexual selection in turtles (order Testudines). Oecologia 44: 185-191.

Berteaux, D. 1993. Female-biased mortality in a sexually dimorphic ungulate: feral cattle of Amsterdam Island. Journal of Mammalogy 74: 732-737.

Björklund, M. 1990. A phylogenetic interpretation of sexual dimorphism in body size and ornament in relation to mating system in birds. Journal of Evolutionary Biology 3: 171-183.

Björklund, M. 1991. Evolution, phylogeny, sexual dimorphism and mating system in the grackles (Quiscalus spp: Icterinae). Evolution 45: 608-621.

Cabana, G., Frewin, A., Peters, R. H. and Randall, L. 1982. The effect of sexual size dimorphism on variations in reproductive effort of birds and mammals. American Naturalist 120: 17-25.

Cheverud, J. M., Dow, M. M. and Leutenegger, W. 1985. The quantitative assessment of phylogenetic constraints in comparative analyses: sexual dimorphism in body weight among primates. Evolution 39: 1335-1351.

Cheverud, J. M., Dow, M. M. and Leutenegger, W. 1986. A phylogenetic autocorrelation analysis of sexual dimorphism in primates. American Anthropologist 88: 916-922.

Clutton-Brock, T. H., Albon, S. D. and Guinness, F. E. 1985. Parental investment and sex differences in juvenile mortality in birds and mammals. Nature 313: 131-133.

Clutton-Brock, T. H. and Harvey, P. H. 1977. Primate ecology and social organization. Journal of Zoology, London 183: 1-39.

Clutton-Brock, T. H., Harvey, P. H. and Rudder, B. 1977. Sexual dimorphism, socionomic sex-ratio and body weight in primates. Nature 269: 797-800.

Coddington, J. A. 1990. Bridges between evolutionary pattern and process. Cladistics 6: 379-386.

Darwin, C. 1859. On the origin of species by means of natural selection. Penguin Books, London.

Darwin, C. 1871. The descent of man and selection in relation to sex. Murray, London.

Downhower, J. F. 1976. Darwin's finches and the evolution of sexual dimorphism in body size. Nature 263: 558-563.

Downhower, J. F., Brown, L., Pederson, R. and Staples, G. 1983. Sexual selection and sexual dimorphism in mottled sculpins. Evolution 37: 96-103.

Eisen, E. J. and Hanrahan, J. P. 1972. Selection for sexual dimorphism in body weight of mice. Australian Journal of Biological Sciences 25: 1015-1024.

Elgar, M. A. 1991. Sexual cannibalism, size dimorphism, and courtship behaviour in orb-weaving spiders (Araneidae). Evolution 45: 444-448.

Elgar, M. A., Ghaffar, N. and Read, A. F. 1990. Sexual dimorphism in leg length among orb-weaving spiders: a possible role for sexual cannibalism. Journal of Zoology, London 222: 455-470.

Ellis, L. 1995. Dominance and reproductive success among nonhuman animals: a cross-species comparison. Ethology and Sociobiology 16: 257-333.

Ely, J. and Kurland, J. A. 1989. Spatial autocorrelation, phylogenetic constraints, and the causes of sexual dimorphism in primates. International Journal of Primatology 19: 151-171.

Erlinge, S. 1979. Adaptive significance of sexual dimorphism in weasels. Oikos 33: 233-245.

Fairbairn, D. J. 1990. Factors influencing sexual size dimorphism in temperate waterstriders. American Naturalist 136: 61-86.

Fairbairn, D. J. and Preziosi, R. F. 1994. Sexual selection and the evolution of allometry for sexual size dimorphism in the waterstrider, Aquarius remigis. American Naturalist 144: 101-118.

Ford, S. M. 1994. Evolution of sexual dimorphism in body weight in Platyrrhines. American Journal of Primatology 34: 221-244.

Forsman, A. 1991. Variation in sexual size dimorphism and maximum body size among adder populations: effects of prey size. Journal of Animal Ecology 60: 253-267.

Gaulin, S. J. C. and Boster, J. S. 1992. Human marriage systems and sexual dimorphism in stature. American Journal of Physical Anthropology 89: 467-475.

Gaulin, S. J. C. and Sailer, L. D. 1984. Sexual dimorphism in weight among the primates: the relative impact of allometry and sexual selection. International Journal of Primatology 5: 515-535.

Gaulin, S. J. C. and Sailer, L. D. 1985. Are females the ecological sex? American Anthropologist 87: 111-119.

Greenwood, P. J. and Adams, J. 1987. Sexual selection, size dimorphism and a fallacy. Oikos 48: 106-108.

Greenwood, P. J. and Wheeler, P. 1985. The evolution of sexual size dimorphism in birds and mammals: a 'hot blooded' hypothesis. Pages 287-299 in P. J. Greenwood, P. H. Harvey and M. Slatkin, eds. Evolution: essays in honour of John Maynard Smith. Cambridge: Cambridge University Press.

Halliday, T. R. 1994. Sex and evolution. Pages 150-192 in P. J. Slater and T. R. Halliday, eds. Behavior and evolution. Cambridge University Press, Cambridge.

Halliday, T. R. and Verrell, P. A. 1986. Sexual selection and body size in amphibians Herpetological Journal 1: 86-92.

Harcourt, A. H., Harvey, P. H., Larson, S. G. and Short, R. V. 1981. Testis weight, body weight and breeding system in primates. Nature 293: 55-57.

Harvey, A. W. 1990. Sexual differences in contemporary selection acting on size in the hermit crab Clibanarius digueti. American Naturalist 136: 292-304.

Harvey, P. H. and Clutton-Brock, T. H. 1985. Life history variation in primates. Evolution 39: 559-581.

Harvey, P. H., Martin, R. D. and Clutton-Brock, T. H. 1987. Life histories in comparative perspective. Pages 181-196 in B. B. Smuts, D. L. Cheney, R. M. Seyfarth, R. W. Wrangham, and T. T. Struhsaker, eds. Primate societies. University of Chicago Press, Chicago.

Höglund, J. 1989. Size and plumage dimorphism in lek-breeding birds: a comparative analysis. American Naturalist 134: 72-87.

Höglund, J. and Sillén-Tullberg, B. 1994. Does lekking promote the evolution of male-biased size dimorphism in birds? On the use of comparative approaches. American Naturalist 144: 881-889.

Howard, R. D. 1981. Sexual dimorphism in bullfrogs. Ecology 62: 303-310.

Houston, D. and Shine, R. 1993. Sexual dimorphism and niche divergence: feeding habits of the Arafura filesnake. Journal of Animal Ecology 62: 737-748.

Jarman, P. 1983. Mating system and sexual dimorphism in large terrestrial, mammalian herbivores. Biological Review 58: 485-520.

Jönsson, P. E. and Alerstam, T. 1990. the adaptive significance of parental role division and sexual size dimorphism in breeding shorebirds. Biological journal of the Linnean Society 41: 301-314.

Kappeler, P. M. 1990. The evolution of sexual size dimorphism in prosimian primates. American Journal of Primatology 21: 201-214.

Kappeler, P. M. 1991. Patterns of sexual dimorphism in body weight among prosimian primates. Folia Primatologica 57: 132-146.

Kozlowski, J. 1989. Sexual size dimorphism: a life history perspective. Oikos 54: 253-255.

Lande, R. 1980. Sexual dimorphism, sexual selection, and adaptation in polygenic characters. Evolution 34: 292-305.

Leutenegger, W. 1978. Scaling of sexual dimorphism in body size and breeding system in primates. Nature 272: 610-611.

Leutenegger, W. 1982. Scaling of sexual dimorphism in body weight and canine size in primates. Folia Primatologica 37: 163-176.

Leutenegger, W. and Cheverud, J. 1982. Correlates of sexual dimorphism in primates: ecological and size variables. International Journal of Primatology 3: 387-402.

Leutenegger, W. and Cheverud, J. M. 1985. Sexual size dimorphism in primates: the effects of size. Pages 33-50 in W. L. Jungers, ed. Size and scaling in primate biology. Plenum Press: New York.

Leutenegger, W. and Lubach, G. 1987. Sexual dimorphism, mating system, and effect of phylogeny in De Brazza's Monkey (Cercopithecus neglectus). American Journal of Primatology 13: 171-179.

Lovich, J. E. and Gibbons, J. W. 1992. A review of techniques for quantifying sexual size dimorphism. Growth, Development & Ageing 56: 269-281.

Madsen, T. and Shine, R. 1994. Costs of reproduction influence the evolution of sexual size dimorphism in snakes. Evolution 48: 1389-1397.

Margulis, L. and Sagan, D. 1984. Evolutionary origins of sex. Pages 16-47 in R. Dawkins and M. Ridley, eds. Oxford Survey in Evolutionary Biology 1.

Margulis, L. and Sagan, D. 1985. Origins of sex: three billion years of genetic recombination. Yale University Press, New Haven.

Martin, R. D. 1980. Sexual dimorphism and the evolution of higher primates. Nature 287: 273-275.

Martin, R. D., Willner, L. A. and Dettling, A. 1994. Pages 159-200 in R. V. Short and E. Balaban, eds. The differences between the sexes. Cambridge University Press: Cambridge.

Maynard Smith, J. 1978. The evolution of sex. Cambridge University Press: Cambridge.

McLachlan, A. 1986. Sexual dimorphism in midges: strategies for flight in the rain-pool dweller Chironomus imicola (Diptera: Chironomidae). Journal of Animal Ecology 55: 261-267.

McLain, D. K. 1987. Heritability of size, a sexually selected character, and the response to sexual selection in a natural population of the southern green stinkbug, Nezara viridula (Hemiptera: Pentatomidae). Heredity 59: 391-395.

Millar, J. S. and Hickling, G. J. 1990. Fasting endurance and the evolution of mammalian body size. Functional Ecology 4: 5-12.

Miller, S. A. and Harley, J. P. 1992. Zoology. Wm. C. Brown Publishers: Dubuque.

Møller, A. P. 1991. Sperm competition, sperm depletion, paternal care, and relative testes size in birds. American Naturalist 137: 882-906.

Mueller, H. C. and Meyer, K. 1985. The evolution of reversed sexual dimorphism in size. A comparative analysis of the falconiformes of the Western Palearctic. Pages 65-101 in R. F. Johnston, ed. Current ornithology volume 2. Plenum Press: New York.

Nylin, S., Wiklund, C., Wickman, P. -O. and Garcia-Barros, E. 1993. Absence of trade-offs between sexual size dimorphism and early male emergence in a butterfly. Ecology 74: 1414-1427.

Nylin, S. and Wedell, N. 1994. Sexual size dimorphism and comparative methods. Pages 253-280 in P. Eggleton and R. I. Vane-Wright, eds. Phylogenetics and ecology. London: Academic Press.

Oakes, E. J. 1992. Lekking and the evolution of sexual dimorphism in birds: comparative approaches. American Naturalist 140: 665-684.

Orians, G. H. 1969. On the evolution of mating systems in birds and mammals. American Naturalist 103: 589-603.

Owen-Smith, N. 1993. Comparative mortality rates of male and female kudus: the costs of sexual size dimorphism. Journal of Animal Ecology 62: 428-440.

Oxnard, C. E. 1983. Sexual dimorphism in the overall proportions of primates. American Journal of Primatology 4: 1-22.

Parker, G. A. 1992. The evolution of sexual size dimorphism in fish. Journal of Fish Biology 41(Supplement B): 1-20.

Parker, G. A., Baker, R. R. and Smith, V. G. F. 1972. The origin and evolution of gamete dimorphism and the male-female phenomenon. Journal of Theoretical Biology 36: 529-553.

Payne, R. B. 1984. Sexual selection, lek and arena behavior, and sexual size dimorphism in birds. Pages 1-52 in M. S. Foster, ed. Ornithological monographs no. 33. Allen Press: Lawrence, Kansas.

Peters, R. H. 1983. The ecological implications of body size. Cambridge University Press: Cambridge.

Pickford, M. 1986. On the origins of body size dimorphism in primates Pages 77-91 in M. Pickford and A. B. Chiarelli, eds. Sexual dimorphism in living and fossil primates. Il Sedicesimo: Firenze.

Pilbeam, D. and Gould, S. J. 1974. Size and scaling in human evolution. Science 186: 892-901.

Price, T. D. 1984a. The evolution of sexual size dimorphism in Darwin's finches. American Naturalist 123: 500-518.

Price, T. D. 1984b. Sexual selection on body size, territory and plumage variables in a population of Darwin's finches. Evolution 38: 327-341.

Promislow, D. E. L., Montgomerie, R. and Martin, T. E. 1992. Mortality costs of sexual dimorphism in birds. Proceedings of the Royal Society in London Series B 250: 143-150.

Ralls, K. 1976. Mammals in which females are larger than males. Quarterly Review of Biology 51: 245-276.

Ranta, E., Laurila, A., Elmberg, J. 1994. Reinventing the wheel: analysis of sexual dimorphism in body size. Oikos 70: 313-321.

Rice, A. R. 1984. Sex chromosomes and the evolution of sexual dimorphism. Evolution 38: 735-742.

Rowell, T. E. and Chism, J. 1986. Sexual dimorphism in mating systems: jumping to conclusions. Pages 107-111 in M. Pickford and B. Chiarelli, eds. Sexual dimorphism in living and fossil primates. Il Sedicesimo: Firenze.

Searcy, W. A. 1979. Sexual selection and body size in male red-winged blackbirds. Evolution 33: 649-661.

Selander, R. K. 1972. Sexual selection and dimorphism in birds. Pages 180-230 in B. Campbell, ed. Sexual selection and the descent of man: 1871-1971. Aldine, Chicago.

Shine, R. 1978. Sexual size dimorphism and male combat in snakes. Oecologia 33: 269-277.

Shine, R. 1989. Ecological causes for the evolution of sexual dimorphism: a review of the evidence. Quarterly Review of Biology 64: 419-461.

Shine, R. 1990. Proximate determinants of sexual differences in adult body size. American Naturalist 135: 278-283.

Shine, R. 1991. Intersexual dietary divergence and the evolution of sexual dimorphism in snakes. American Naturalist 138: 103-122.

Singer, M. C. 1982. Sexual selection for small size in small butterflies. American Naturalist 119: 440-443.

Slatkin, M. 1984. Ecological causes of sexual dimorphism. Evolution 38: 622-630.

Smith, R. J. 1993. Logarithmic transformation bias in allometry. American Journal of Physical Anthropology 90: 215-228.

Speakman, J. R. 1992. Evolution of animal body size: a cautionary not on assessments of the role of energetics. Functional Ecology 6: 495-498.

Stamps, J. A. 1993. Sexual size dimorphism in species with asymptotic growth after maturity. Biological Journal of the Linnean Society 50: 123-145.

Stamps, J. A., Krishnan, V. V. and Andrews, R. M. 1994. Analyses of sexual size dimorphism using null growth-based models. Copeia 3: 598-613.

Stearns, S. C. 1987. Why sex evolved and what difference it makes. Pages 15-31 in S. C. Stearns, ed. The evolution of sex and its consequences.

Stearns, S. C. 1992. The evolution of life histories. Oxford University Press: Oxford.

Steele, R. H. and Partridge, L. 1988. A courtship advantage for small males in Drosophila subobscura. Animal Behaviour 36: 1190-1197.

Sullivan, B. K. 1984. Size dimorphism in anurans: a comment. American Naturalist 123: 721-724.

Svärd, L. and Wiklund, C. 1989. Mass and production rate of ejaculates in relation to monandry/polyandry in butterflies. Behavioural Ecology and Sociobiology 24: 395-402.

Teather, K. L. and Weatherhead, P. J. 1994. Allometry, Adaptation, and the growth and development of sexually dimorphic birds. Oikos 71: 515-525.

Trail, P. W. 1990. Why should lek-breeders be monomorphic? Evolution 44: 1837-1852.

Trivers, R. L. 1972 Parental investment and sexual selection. Pages 136-179 in B. Campbell, ed. Sexual selection and the descent of man, Aldine Publishing Company: Chicago.

Trivers, R. L. 1985. Social evolution. Benjamin/Cummings, Menlo Park, California.

Vollrath, F. and Parker, G. A. 1992. Sexual dimorphism and distorted sex ratios in spiders. Nature 360: 156-159.

Ward, P. I. 1987. Sexual selection and body size in Gammarus pulex: a reply to Greenwood and Adams. Oikos 48: 108-109.

Weatherhead, P. J. and Clark, R. G. 1994. Natural selection and sexual size dimorphism in red-winged blackbirds. Evolution 48: 1071-1079.

Wedell, N. 1993. Spermatophore size in bushcrickets: comparative evidence for nuptial gifts as sperm competition device. Evolution 47: 1203-1212.

Wiklund, C. and Fagerström, T. 1977. Why do males emerge before females? A hypothesis to explain the incidence of protandry in butterflies. Oecologia 31: 153-158.

Wiklund, C. and Forsberg, J. 1991. Sexual size dimorphism in relation to female polygamy and protandry in butterflies: a comparative study of Swedish Pieridae and Satyridae. Oikos 60: 373-381.

Wiklund, C. and Karlsson, B. 1988. Sexual size dimorphism in relation to fecundity in some Swedish Satyrid Butterflies. American Naturalist 131: 132-138.

Wiklund, C., Nylin , S. and Forsberg, J. 1991. Sex-related variation in growth rate as a result of selection for large size and protandry in a bivoltine butterfly, Pieris napi. Oikos 60: 241-250.

Willner, L. A. and Martin, R. D. 1985, Some basic principles of mammalian sexual dimorphism. Pages 1-42 in J. Ghesquire, R. D. Martin and F. Newcombe, eds. Human sexual dimorphism. Taylor and Francis: London.

Woolbright, L. L. 1983. Sexual selection and size dimorphism in anuran amphibia. American Naturalist 121: 110-119.