Polyandry and allele frequency differences between the sexes in the ant Formica aquilonia

Abstract

Genetic mother-offspring analyses based on six enzyme gene loci show that about 60 per cent of the females of the mound-building red wood ant Formica aquilonia mate with several males. The number of matings inferred from the offspring genotypes range from one to six, the arithmetic mean being 1.94. The mates do not contribute equally in the inseminations; in the case of two matings, one male is estimated to inseminate on average 77 per cent of the offspring. The average relatedness among the offspring of a single female is 0.6, corresponding to the effective number of matings of 1.43. Three of the six loci show remarkable allele frequency differences between the sexes. At each of these loci the males virtually lack the alleles present in females with frequencies 0.18 at Me, 0.25 at Pgk and 0.27 at Gpi. Segregation analyses indicate normal Mendelian inheritance at these loci and the difference between the sexes seem likely to result from selection.

Multiple mating and facultative polygyny in the Panamanian leafcutter ant Acromyrmex echinatior

“…The non-detection error of patrilines due to limited variation at the marker locus can be quanti®ed precisely only for populations with single and double queen mating (Pamilo 1982;Pedersen and Boomsma, in press). However, as a crude approximation, one may use the sum of the squared populationwide allele frequencies among males as an average probability for each observed patriline to have been two patrilines carrying an identical father allele (Pamilo 1993;Boomsma and Ratnieks 1996). This average non-detection error was between 26 and 41%, depending on whether the observed allele frequencies among putative fathers or queens were used, respectively, implying that, on average, roughly one queen mate with a signi®cant share in worker paternity was missed in each colony.…”

“…If the queen's mates do not contribute equally to her o spring (paternity skew), the e ective mating frequency is lower than the absolute number of males and average relatedness between colony o spring is higher (Starr 1979(Starr , 1984Boomsma and Ratnieks 1996). E ective queen mating frequency was estimated for each colony following Pamilo (1993), ®rstly as the harmonic-mean mating frequency m eYp 1aRp 2 i , where p i is the proportional contribution of male i) and secondly corrected for binomial sampling error {m eYy 1a x Rp 2 i À 1 a x À 1 , where N is the number of o spring analysed for each individual colony}. Assuming random mating, and following the notation of Pamilo (1993), the relatedness (g) among nestmates then follows from the relationship g 0.25 + 1/2m e .…”

“…The mating frequency of queens (expressed as the number of putative patrilines observed within a colony) and the parental genotypes were inferred from the observed genotypes of queens and their o spring (Pamilo 1993). When queen genotypes were not available (due to non-destructive sampling), these could be reconstructed, either from worker genotypes (n 4) or on the basis of colony male o spring (n 5) (assuming that workers do not normally lay viable eggs themselves; see Results for a justi®cation of this assumption).…”

“…E ective queen mating frequency was estimated for each colony following Pamilo (1993), ®rstly as the harmonic-mean mating frequency m eYp 1aRp 2 i , where p i is the proportional contribution of male i) and secondly corrected for binomial sampling error {m eYy 1a x Rp 2 i À 1 a x À 1 , where N is the number of o spring analysed for each individual colony}. Assuming random mating, and following the notation of Pamilo (1993), the relatedness (g) among nestmates then follows from the relationship g 0.25 + 1/2m e . This equation gives colony-level relatedness estimates based on pedigree information, inferred via parent-o spring analysis.…”

“…A considerable number of ant species have been reported to mate multiply, but most of these reports have been observations of multiple copulations, whereas documented cases of high e ective queen mating frequencies in ants are rare (see Boomsma and Ratnieks 1996 for a review). Ant species or populations where both polyandry and polygyny signi®cantly reduce relatedness among nestmate workers have so far only been reported in Formica (Pamilo 1993;SundstroÈ m 1993).…”