A Bruce Effect in Wild Geladas

Science  09 Mar 2012:
Vol. 335, Issue 6073, pp. 1222-1225
DOI: 10.1126/science.1213600
  • Fig. 1

    Number of births for groups with and without male replacements. Sum of births from groups with a male replacement (light gray) as compared with a similar-sized group at the same time that did not experience a male replacement (black). Paired groups were compared (Wilcoxon signed-ranks test) 0 to 6 months before male replacement, 0 to 6 months after male replacement, and 7 to 12 months after male replacement.

  • Fig. 2

    (A) Hormonally determined pregnancy failures. Hormone (fE ng/g) profiles of three representative females that aborted after male replacement. Each black diamond represents one hormone sample from a given female. Dark vertical lines [for both (A) and (B)] indicate the time of male replacement. (B) Composite comparison between hormone (fE ng/g) profiles of successful (“successful,” black circles) and aborted (“failed,” gray squares) pregnancies across each trimester (±SEM). Estrogens of successful and aborted pregnancies show no significant differences before male replacement. After male replacement (vertical line), aborted pregnancy estrogen profiles drop well below pregnancy levels. Sample size (number of pregnancies) is indicated in italics above the x axis. One female that aborted was not included in the figure because we were unable to obtain a hormone sample from her immediately after male replacement; she is known to have aborted because she expelled a poorly developed fetus (23) the day after a male replacement.

  • Fig. 3

    Boxplot of interbirth intervals between successfully weaned infants (whiskers represent highest and lowest values). Interbirth intervals (in years) for females experiencing different events between infants. None: two successive infants; pregnancy termination: an abortion between surviving infants; infant death: an infant death between surviving infants. Sample size (number of interbirth intervals) is indicated in italics above the x axis.

  • Table 1

    Fecal hormone sample schedule and collection rates.

    GroupsFemalesFecal samples*Sample rate
    High coverage5103311 sample/1.9 days
    Mid coverage14626701 sample/7.6 days
    Low coverage188810001 sample/29.2 days

    *Number of females in each category does not include all females in the group. High-coverage females were selected based on probability of being or becoming pregnant.

    Groups and females are represented by more than one “coverage” schedule; therefore, the total number of groups and females does not add up to the sum of each category.

    Additional Files