In species with biparental care and lifetime monogamy, the fecundity of a male’s partner can be a major component of his fitness but it is unclear whether males can assess female fecundity before breeding. We carried out an experiment in which we measured variation in female fecundity (repeatability 39%, 213 females) in a captive zebra finch population and tested whether males preferred unfamiliar females of high fecundity (approximately top 10% of the population; 30 eggs laid on average) over those of low fecundity (bottom 10%; 6 eggs). We first tested whether naïve human observers could identify the high-fecundity female when confronted with duos of high and low fecundity. Humans guessed correctly in 58% of the cases (95% confidence interval [CI] 50–66%) indicating that differences in female condition were not highly obvious to humans. Zebra finch males preferred the high-fecundity female in 59% of choice tests that lasted 20 min (CI 52–66%). When extending such choice tests over several days, male “success” in associating with the high-fecundity female was still modest (61% correct choices, CI 44–76%). Overall, male zebra finches seem to have only limited abilities to identify the better mate when faced with a choice between extremes in terms of female fecundity. We found no male preference for heavier females. We speculate that such a preference may not have evolved because, in contrast to many ectothermic species, predicting fecundity from female weight is not sufficiently accurate (r2 = 0.04) for the benefits to outweigh the costs of increased male–male competition for heavy females.
fecundity_data_of_all_females
This file includes fecundity variation of all the females from which 20 high-fecundity and low-fecundity duos were selected (Table S3). The linear mixed models describe variation in female fecundity (i.e. the number of eggs laid per 7-week breeding round; i.e. ‘eggs’; this is ‘estimated fecundity’ before genetic parentage assignment because females were chosen based on that) as a function of female identity (FID), breeding round (first vs. second 7-week period), the number of days females were socially paired (‘dayspaired’), and the exclusiveness of the pair bond (‘Fshare’, with high values reflecting exclusive pairing). The column labelled ‘F’ refers to the inbreeding coefficient of each female.
summary_of_20_femaleduos
This file includes all information on the selected 20 duos (Table 1), where ‘duoIDwithin’ is the ID of the female duo within each replicate, and where ‘high’ and ‘low’ always refer to traits of the high-fecundity and low-fecundity females in that duo, respectively. In this file, ‘higheggssoc’ and ‘loweggssoc’ refer to ‘estimated fecundity’ in Table 1, while ‘highGeggs&IF’ and ‘lowGeggs&IF’ mean ‘true fecundity’ in that Table (Table 1). This file also includes all morphological trait values (beak color score: ‘BChigh’, ‘BClow’; mass (g): ‘masshigh’, ’masslow’), age (days) and female responsiveness during the test (‘reshigh’, ‘reslow’) of these 20 duos of females. We also included the inbreeding coefficient of these females (‘F_high’, ‘F_low’).
480_male_10min_tests
This file gives raw data from two-way choice tests: 240 trials of 10min per replicate. After 10min of trial sides are switched, so each ‘Supertest’ of 20 min (see next file #4) consists of two consecutive 10min trials (‘Round’). ‘duoIDwithin’ is the ID of each female duo within replicates. ‘L’ and ‘R’ always refer to the left and right side of the choice chamber, respectively, hence ‘MID’, ‘FID_L’, and ‘FID_R’ are the IDs of the male, the female on the left side, and the female on the right side of the cage. ‘SingL’ is seconds of song to the female on the left. ‘Attention’ is seconds of paying attention without singing to the female and ‘Neutral’ is time not paying attention or in the center. ‘RespL-R’ is the responsiveness of the left minus the right female, ‘PropAtt&SingL’ is the proportion of time paying attention or singing to the left female, ‘FecundFemaleL’ identifies the female on the left as either high or low fecundity; hence, ‘PropAtt&SingHihFec’ is then the attention given to the high-fecundity female. Cases of missing data are explained in ‘Comments’.
227_male_20min_tests
This file gives data for the 227 out of 240 planned ‘Supertests’ of 20min (after excluding 13 tests cancelled because two females died). This is the data for Figure1, Table 2 and Table S4. Here ‘high’ and ‘low’ always refer to traits of the high- and low-fecundity females, respectively, and ‘difference’ always refers to ‘high’ minus ‘low’. Given are the individual IDs, age in days, beak colour (‘BC’), body mass, and responsiveness. ‘Maletest’ refers to a male’s first vs. second choice test, ‘male_choosing_motivation’ is the sum of seconds the male is actively paying attention (theoretical max = 1,200 sec), ‘PrefHighFec’ is then the relative time with the high-fecundity female and ‘correct’ is 1 when ‘PrefHighFec’>0.5. Cases of missing data are explained in ‘comment’.
human_rating
This file includes the original data of the ‘Human rating of female fecundity’ experiment. ‘observerID” refers to the 77 participants of the study, ‘order’ refers to the order in which they rated the duos of females, ‘duoIDwithin’ is the ID of each female duo within replicates, and ‘correct’ is the outcome of whether the high-fecundity female was correctly identified by the observer (0=no, 1=yes).
nest_building
This file includes original data of the ‘Two-way choice tests: nest-building’ experiment. ‘duoIDwithin’ is the ID of each female duo within replicates, ‘MID’ is the ID of the choosing male, and ‘correct’ is the outcome of whether the male preferred the high-fecundity female (0=no, 1=yes).