• Van Gossum, Hans
  • Robb, Tonia
  • Forbes, Mark R.
  • Rasmussen, L.


In several animal species, one male type coexists with two to several female types, a polymorphism often explained in the context of sexual selection. Where it occurs, one female morph typically resembles the conspecific male phenotype, but the degree of resemblance varies across species. Here, we question whether the degree of phenotypic similarity between male-like females and males varies within species. Phenotypic resemblance is hypothesized to depend on the potential for frequency- and density-dependent selection on male and (or) female phenotypes. We studied six populations of the damselfly Nehalennia irene (Hagen, 1861) that differed widely in estimates of morph frequency and male density. Male-like females resemble males more than another female type resembles males, across populations, when comparisons are based on abdominal patterns. Abdomen phenotype does matter in male–female interactions of damselflies. Furthermore, male-like females were more similar to males at low and high density sites compared with sites with intermediate densities, contrary to the hypothesis that the potential for male harassment influences the degree of phenotypic similarity. Additionally, male-like females of most populations converged on the abdominal phenotype of males of one population rather than on that of syntopic males; a problem that has not received any attention.


Nehalennia irene is a small (body length 25–28 mm; Cannings and Stuart 1977) and slender damselfly that is distributed across Canada, excluding the far north (Walker 1953). For all studied populations of this species, two female morphs coexist (see Forbes et al. 1995; Van Gossum et al. 2007a). One female morph is green and bluish on the thorax and has large paired dark spots on the dorsum of segment IX, similar to males (Fig. 1a–1f) (hereafter andromorph or gynomorph). The other female morph (hereafter heteromorph) is green and yellowish on the thorax and lacks the distinctive triangular patch which andromorphs show on segment VIII (Fig. 1g–1i; for further detailed description see Forbes et al. 1995 and for colour drawings see Lam 2004). Males, andromorphs, and gynomorphs all have the pale parts on the abdomen coloured blue, and for this body part only, differ in the amount of black patterning. While there is continuous variation in patterning within males and andromorphs and within heteromorphs, this patterning is distinct between morphs (see Results). Nehalennia irene frequents marshy or boggy areas with relatively dense vegetation, where it flies close to the water among the rushes or other emergent plants, and in the grass and sedge along the shore, but it usually does not venture over open water (Walker 1953). It is probably the most abundant damselfly in eastern Canada and is by no means rare in the western Canadian provinces (Walker 1953). This species is active from the end of May and extends into August (e.g., Whitehouse 1941; Walker 1953; Lajeunesse and Forbes 2003; Van Gossum et al. 2007a).

We chose N. irene because of the characteristics of the abdomen phenotype in males and female morphs, and because it shows wide spatial variation in female morph frequencies and male densities (see Van Gossum et al. 2007a). Although populations in eastern and western Canada are consistently low in male density (9.22 ± 5.96 (mean ± SD) individuals caught/min of standardized netting in Western Canada and 4.34 ± 4.18 males caught/min of standardized netting in Eastern Canada; for details see Van Gossum et al. 2007a), those in the west show high frequencies of andromorphic females (0.95 ± 0.03), while those in the east show low frequencies of andromorphic females (0.03 ± 0.07). For central Canadian populations, the frequencies of andromorphic females vary from 0 to ~45% and male densities range from low to high (19.83 ± 12.61 individuals/min).

Frequencies and densities of male and female morphs were estimated for six different populations across Canada between 16 June and 9 July 2004 (Table 1). All populations were >5 km apart and were not part of the same lake or river. Damselflies were collected by sweeping a “figure eight” with an insect net through the vegetation at the water’s edge (where mating occurs). Five to 10 sweeps were made per site and for each site all sampling occurred on 1 day only. Sex and female morph was recorded for each individual collected, and all individuals were marked prior to their release to prevent multiple counts of an individual. Female morph frequencies were calculated as the ratio of andromorphs (i.e., number of andromorphs to total number of females). To estimate male density, the number of male damselflies caught was expressed per unit of time spent sampling (several sweeps were completed at each population). Chi-square tests of homogeneity between populations were completed to determine if there was significant variation in densities and female morph frequencies. Pairwise analyses with Holm’s correction factor were used to determine populations that were significantly different.