You're reading...
My Research

Body Length Frequency Distributions Among Male and Female Gammarus minus

Authors: S. Lee Ware and Krystina Sandefur



specimen from gammarus genus

Two aspects of the Gammarus minus species were examined in this study, (1) sexual dimorphism and (2) mate choice as a function of body length (mm). Both amplexus and non-amplexus specimens in a wild population sample of Gammarus minus were measured from the base of the antenna to tip of the uropod. Amplexing male mean length was observed to be 8.62 ± 1.17 mm, paired with an amplexed female mean length of 5.62 ± 0.9 mm. The non-amplexing males and non-amplexed females were observed to have a mean length of 6.86 ± 1.79 mm and 5.27 ± 1.14 mm, respectively. The body length differences between males and females as well as between amplexing males and non-amplexing males was found to be significant, while the body length difference between amplexed females and non-amplexed females was found to be insignificant. Twenty-one percent of precopula male-female pairs displayed a size ratio of 1.5 : 1, with 91% displaying a size ratio of 1.3 : 1 or greater. These results confirm sexual dimorphism and a significant probability of size-related sexual selection among G. minus.


Gammarus minus is one of approximately 200 species of the genus Gammarus – amphipod crustaceans widely distributed throughout the Holarctic and Northern Hemisphere tropics (Gammarus, 2009). Amphipods are generally small with laterally compressed bodies, seven thoracic segments, therefore seven pairs of thoracic legs, a six-segmented abdomen, and a small tail. Most are active at night and bottom-scavenge for food. Gammarus minus breed year-round, can be found in abundant numbers in spring-fed waters, and are easily maintained in laboratory settings, making them a convenient organism to study (Glazier, 2000).

Sexual dimorphism is a common phenomenon throughout the animal kingdom. Differences in size, color, behavior, and ornament between sexes of the same species have been attributed to various evolutionary processes such as the handicap principle and polyamory (Sexual Dimorphism, 2009). However, contrary to the overall pattern of females being larger than males, the sexual size dimorphism (SSD) in amphipods is reversed: males are larger than females. The strongest candidate hypothesis to explain this phenomenon is sexual selection.

Sexual selection is a form of intraspecific competition (competition within a species), where either (or both) intrasexual competition for mate access or intersexual selection determines the reproductive success of individuals of that species (Sexual Selection, 2009).
Sexual selection appears to be a driving force in amphipod evolution, affecting phenotypic characteristics such as behavior (reproductive strategies, id est mate guarding) and body size (sexual dimorphism).

During breeding, the Gammarus male carries the female as a precopula pair until the female molts, at which time the male fertilizes her eggs. The fertilized eggs are carried in a marsupium brood pouch until hatching. As stated, G. minus males are larger than females, and in studies of British Gammarus species, males have been shown to compete intrasexually for access to females where larger body size results in a greater number of successful matings (Glazier, 2000). However, anatomical, environmental, or other constraints have kept overall Gammarus body sizes relatively low. For example, one significant constraint seems to be inherent in the precopula ritual – males must be significantly larger than their female mates for successful amplexus, but the larger the female, the more offspring that are produced (Clarkson & Birkhead, 1980). Larger size would thus seem to add evolutionary incentive for both males and females. However, larger precopula pairs are more conspicuous to predators and less able to evade them (Ward, 1986), contributing to a determinate constraint on overall Gammarus body sizes.

This study tests three basic hypotheses: (1) if larger size is advantageous in mate guarding, adult males in precopula should be larger, on average, than adult males not in precopula. (2) since sexual selection is expected to be acting on males only, there should be no significant difference in body size between amplexed females and non-amplexed females. (3) since males must be significantly larger than their female mates for successful amplexus, one would expect to find preferential male / female body length ratios in precopula pairs.


The wild Gammarus minus population sampled in this study was collected at a small spring emerging from Picnic Cave on highway 25S near Mt. Vernon, Kentucky in the Appalachia region of the eastern United States. The amphipods were collected by sweeping a dipnet across the mossy stream bottom, being careful to avoid disturbing sediment. The specimens, once collected, were then deposited into an enamel pan by inverting the dipnet. The non-amplexing males and non-amplexed females were poured into a cooler filled with stream water, while the precopula pairs were individually pipetted into small jars of spring water.

Before being measured, non-amplexus amphipods were randomly selected with a dipnet after being mixed in an enamel pan. They were then stored individually in plastic containers with stream water in an environmental chamber at 15 °C and a 12 h light : 12 h dark cycle. Beginning with the precopula pairs, the gammarids were sexed (males were larger and mounted atop the females) and placed in a petri dish filled with carbonated water (to separate and anesthetize them) with cm rulers fixed to the bottom of each (to facilitate length measurement). Each gammarid was anchored with insect pins, stretched, and measured from the base of the antenna to the tip of the uropod to determine its length (in millimeters). The body length ratio between the male and female of each precopula pair was also recorded.

While the precopula pairs were easily sexed (the males being larger and mounted atop the females), the non-amplexus amphipods were sexed using a dissecting microscope by placing each specimen on a slide with carbonated water and a cover slip. Males were determined by observation of large gnathopods (nearly twice the size of females) and angularity of head shape, while females were determined by observation of smaller gnathopods, rounded heads, marsupial plates, and often dark ventral patches in the marsupium.

Once all measurements of the sample G. minus population had been taken and all experimental data had been recorded, several statistical tests were conducted. A Student’s t-test for independent samples was used to compare the mean body lengths of: (1) amplexing males and non-amplexing males and (2) amplexed females and non-amplexed females. A Chi-Square “contingency test” was used to compare body length distribution among: (1) amplexing males and non-amplexing males and (2) amplexed females and non-amplexed females. Finally, a Chi-Square Goodness of Fit test was employed to test the body length ratios of precopula males / females. A probability factor of P ≤ .05 was considered significant for all statistical tests.


The observed sampling from a wild G. minus population displayed significant sexual size dimorphism (SSD) between males and females (Figure 1). Female body length varied from 4 to 8 mm for a range of 4 mm (n = 121), while male body length varied from 4 to 11 mm for a range of 7 mm (n = 105; Figure 1). Within these ranges, the mean male body length (7.88 mm ± 1.69) was 44.6 % (2.43 mm) longer than the mean female body length (5.45 mm ± 1.03).

graph of dimorphism results between males and females

Evidence of sexual dimorphism in body length frequency distribution between males and females

The amplexing male’s mean length was 8.623 ± 1.17 mm (N = 61), while the non-amplexing male’s mean length was 6.864 ± 1.79 mm (N = 44), which were determined to be significantly different (t = 6.065; df = 103). The body length frequency distributions of amplexing vs. non-amplexing male amphipods are summarized in Figure 2. Approximately 63.8 % of the observed male population of G. minus measured between 8 and 10 mm in length; 78 % of this majority were amplexing at the time of capture, while only 34 % of the observed male population measured ≤ 7 mm, with 78 % of this minority not amplexing at the time of capture. Eighty-five percent of the amplexing males were between 8 and 10 mm, while only 34.1 % of the non-amplexing males fell between these lengths. Sixty-four percent of the non-amplexing males vs. only 13 % of the amplexing males were < 8 mm. The Chi Square contingency test confirmed a significant difference in the proportions of the body length distributions between amplexing and non-amplexing males (X2 = 37.75; df = 7).

graph of frequency distributions between amplexing and non-amplexing males

Body length frequency distributions between amplexing vs. non-amplexing males

The amplexed female’s mean length was 5.62 mm ± 0.9 (N = 61), while the non-amplexed female’s mean length was 5.27 mm ± 1.14 (N = 60), which were determined to be insignificantly different (t = 1.826; df = 119). The body length distributions of amplexed vs. non-amplexed female amphipods are summarized in Figure 3. Eighty-six percent of the observed female population of Gammarus minus measured < 7 mm in length. Gammarus minus females measuring between 5 and 6 mm constituted 66 % of the observed female population, with 60 % of this majority being amplexed at the time of capture. Of those females < 5 mm in length, 75 % were not amplexed at the time of capture. Females > 6 mm in length constituted a small minority of the female population and were approximately equal in probability of amplexus. Most significantly divided, 45.9 % of the amplexed females vs. only 21.7 % of non-amplexed females were 6 mm in length, while 30.0 % of the non-amplexed females vs. only 9.8 % of the amplexed females were 4 mm in length (Figure 3). The Chi Square contingency test confirmed the lack of statistical difference in the proportion of body length distributions among both amplexed and non-amplexed females (X2 = 9.3; df = 4).

graph of frequency distributions between amplexing and non-amplexing females

Body length frequency distributions between amplexing vs. non-amplexing females

The male / female amplexus pair body length ratios are summarized in Figure 4. The ratio mean was 1.5 ± 0.262. Ninety-two percent of amplexus pairs had a body length ratio of 1.3 or greater, while 21.3 % of amplexus pairs had a size ratio of 1.5, which was the highest percentage of any ratio group. Males mated almost exclusively with females smaller than themselves, with a possible preference for the 1.5 ratio. A Chi-Square Goodness of Fit test confirmed significant differences in the male / female body length ratio frequency distribution (X2 = 25.93; df = 10).

figure of body length ratios in precopulate pairs

Male to female body length ratio of amplexing precopulate pairs


The experimental data seems to confirm the first of our hypotheses: males in precopula were observed to be larger, on average, than males not in precopula. However, mate guarding was not demonstrated to be the cause of this difference. Other selection factors have been shown to be significant in studies of closely related Gammarus species. For example, results from a set of experiments on Gammarus pulex showed that the swimming performance of sexually dimorphic pairs was consistent with the hypothesis that males are larger as a result of a mechanical constraint and not intrasexual competition for mates (Adams & Greenwood, 1983). Follow-up studies supported the mechanical constraint hypothesis, finding the degree of sexual size dimorphism (SSD) in precopula pairs to be markedly larger for those pairs formed under stream conditions than for those formed in a static water tank (Greenwood & Adams, 1984).

The experimental data also supports our second hypothesis: there was no significant difference in body length between amplexed females and non-amplexed females. This seems surprising at first, because males should preferentially mate with as large a female as possible since she will produce more eggs per brood than a smaller female (Ward, 1988), and since males mating with the largest females would sire more than twice as many offspring as males mating with the smallest females (Clarkson & Birkhead, 1980). So there must be some limiting factor keeping females both smaller than males and more uniform in size (females in our study had a length range of only 4 mm compared to 7 mm among males). This key factor in stabilizing the observed SSD appears to be the body size ratio between males and females in precopula pairs. Ninety-two percent of precopula pairs had a body length ratio of 1.3 or greater (mean of 1.5 ± 0.262), suggesting that only a small minority of G. minus males couple with females close to their own size. One would expect at least 69 % of amplexed females to fall within a standard deviation of the mean body length ratio of the most successful amplexing males (85 % of amplexing males are between 8 and 10 mm). Indeed, that appears to be the case, with 78.7 % of amplexed females measuring between 5 and 6 mm. This suggests that broader mechanisms of natural selection (beyond intrasexual competition) are contributing to the SSD of G. minus. Males are apparently constrained to select only those females smaller than themselves that can be easily amplexed during precopula (Ward, 1988). This limiting ratio factor has been connected to stream conditions as the selective environmental agent in several other studies of closely related species. One such study observed that when the male in an amplexus pair is relatively larger than the female, their swimming performance is superior to that of amplexus pairs in which the male and female are of similar sizes. Ostensibly, this minimizes the risk of being washed downstream by the current (Adams & Greenwood, 1983). Further, since a precopula pair is a much larger and slower moving unit than a single animal, large pairs are more conspicuous to predators and less able to maneuver to avoid them (Ward, 1986).

Unfortunately, there were a number of shortcomings in our plan and procedure. First, our collection techniques did not seek or achieve a random sampling of the wild Gammarus minus population. We specifically sought out a certain number of amplexing pairs, and a certain number of non-amplexus males and females. This makes a determination of the overall condition of the population unknowable. What percentage of the overall population is in amplexus at a given time? What are the overall average lengths of males and females? Future studies could be improved by including overall population estimates coupled with truly random field capture methods. Secondly, our body length measurements were conducted by many different investigators in a classroom setting and are not likely to be reliably accurate or consistent. Therefore, a better measurement is body mass (g), which would be a more accurate measure of body size and would result in more consistent measurements by use of a digital scale, though this method may have its own inherent challenges and limitations.

Based on the results of our study, we see many opportunities for future research. One important issue that is as yet unclear is what other sexual and environmental factors may be contributing to the sexual dimorphism among Gammarus species. Mate guarding and stream conditions are likely contributing factors, but are there others? And to what degree are mate guarding and stream conditions complementing or conflicting with each other in shaping this dimorphism? Another avenue for research should lead to a clearer understanding of what selective forces are limiting the overall increase in G. minus size. Certainly predation is a factor, but to what degree? We believe that successful research raises more questions than it answers, and in this, at least, we feel successful.


The authors would like to thank Ron Rosen for designing this experiment and assigning it to us, and for his excellent feedback during the composition and compilation of this paper. We would also like to thank our Experimental Zoology classmates who worked with us in collecting and measuring the Gammarus minus specimens.


Adams, J., & Greenwood, P. J. (1983). Why are males bigger than females in precopula pairs of Gammarus pulex? Behavioral Ecology and Sociobiology
, 13 (4), 239-241.

Clarkson, K., & Birkhead, T. R. (1980). Mate Selection and Precopulatory Guarding in Gammarus pulex. Z. Tierpsychol
, 52, 365-379.

Gammarus. (2009, February 24). Retrieved April 19, 2009, from Wikipedia.org: http://en.wikipedia.org/wiki/Gammarus

Glazier, D. S. (2000). Sexual Selection and Gammarus minus (amphipoda). Research Link.

Greenwood, P. J., & Adams, J. (1984). Sexual Dimorphism in Gammarus pulex: The Effect of Current Flow on Precopula Pair Formation. Freshwater Biology
, 14, 203-209.

Sexual Dimorphism. (2009, April 9). Retrieved April 10, 2009, from Wikipedia.org: http://en.wikipedia.org/wiki/Sexual_dimorphism

Sexual Selection. (2009, April 3). Retrieved April 10, 2009, from Wikipedia.org: http://en.wikipedia.org/wiki/Sexual_selection

Ward, P. (1986). A Comparative Field Study of Breeding Behavior of a Stream and a Pond Population of Gammarus pulex (Amphipoda). OIKOS
, 46, 29-36.

Ward, P. (1988). Sexual Selection, Natural Selection and Body Size in Gammarus pulex (Amphipoda). The American Naturalist
, 131, 348-349.



Comments are closed.