Female–female cooperation in polygynous oystercatchers

Abstract

Methods Breeding unit size. Donor number is related to the size of the breeding unit (N individuals) as d o Np, where d is the number of different pollen parent genotypes represented in a maternal tree's fruit crop and p is the expected proportion of trees in staminate phase at any given time. Assuming, conserva- tively, that individual trees reproduce asynchronously twice per year 6 and that staminate and pistillate flowering phases each last seven days 20 , p is calculated to be 0.0712. Although these estimates could be effected by mosaicism 21 (the fusion of genetically distinct individuals), its frequency in the species examined here is low. Breeding unit area and radii. These parameters were calculated from paternity-analysis-based estimates of breeding unit size (Nd; Table 2) and the censused densities of adult, reproductively mature trees over 15 km 2 BCI. Based on the long-term censuses of C. Handley and E. Kalko, 6, 108 and 20 adults of F. dugandii, F. obtusifolia and F. popenoei, respectively, are known to occur on BCI. Because these species exhibit little spatial aggregation over the area censused, these densities are assumed to be representative of forested areas surrounding BCI, a conservative assumption given that approximately one-third of the area within 10 km of BCI is occupied by Lake Gatun where figs are absent. Breeding unit radii estimate the distances pollen-bearing, female fig wasps routinely disperse in search of receptive host trees. Although actual breeding populations of figs may deviate substantially from the assumed circular distribution, alternative structures (elliptical, for example) increase estimated wasp dispersal distances.