Brown et al (20) tested female choice based on the peak frequency of calling song in the black-horned tree cricket, O. nigricornis. They began by testing for correlations between frequency and phenotypic correlates of male quality and viability. They found that frequency was a reliable indicator of male size, as demonstrated by its high consistency within males and its correlation with overall body size. Toms (121) tested whether frequency variability in another tree cricket, Oecanthus neosimilis, was due to the rate at which file teeth strike the scraper during song production but found no correlation between frequency and either the total number or density of file teeth. Thus, the relationship between size and frequency is probably not due to differences in the length of the file.
Females preferred synthetic calls with a peak frequency corresponding to that of relatively large males over the higher frequency calls corresponding to smaller males, but only if they had the opportunity to choose between calls simultaneously. Females given individual calls were highly phonotactic, but their degree of phonotaxis did not vary with frequency. Walker (126) showed that female tree crickets remained phonotactic even when frequencies were un- natural, as long as pulse rates were correct. Female preference for intraspecific variation in calling song is therefore not produced by a fixed threshold deci- sion whereby females choose only males above some fixed standard (see 61). Rather, female choice appears to be based on the relative quality of calls that can be sampled simultaneously. Females that mate with large males have higher egg production (see below) (14, 15, 20). Thus song, because of its relationship with size, is a reliable indicator of male quality, and female preference for lower frequency songs appears to be an example of choice for direct benefits. [1]