Several samples of pine forest ants at a body-size locus were compared with the same Hardy-Weinberg expectation at one locus. Which observed-versus-expected comparison best supports departure from Hardy-Weinberg equilibrium?