Sex Ratio Studies

David Astin, Jayme Eggum, and Pete Jamrogiewiez*
Jane Borland, Trey Crumpton and Wendy Allen**
Carol Johnson, Markisha Thomas and Isaac Jemal°
Sonia Altizer, Karen Oberhauser, Michelle Prysby, Liz Goehring and Michelle Solensky°°

*Wayzata HS in Wayzata, MN
**Lamar HS in Arlington, TX
°John Jay HS in San Antonio, TX
°°University of Minnesota, St. Paul MN

We know from studying captive monarchs, and from collecting and raising larvae from the wild, that females lay approximately equal numbers of male and female offspring. However, males and females may behave in ways that result in more of one sex present in a particular place at a particular time. Alternatively, if one sex had a shorter lifespan than the other, the result would be that more individuals of the longer-lived sex would be present at most times. Either of these could result in a biased population sex ratio. In fact, we often collect more of one sex than the other. Assuming that we are equally able to capture males and females, a biased capture sex ratio indicates that the population of monarchs from which we are sampling also has a biased sex ratio. We can use sex ratio data in studies of differences in male and female behavior.

Statistical Note: Even if a population of butterflies has equal numbers of males and females, it is unlikely that we would capture exactly the same number of both sexes when we sampled them. We thus need to use a statistical test to determine if a difference we observe actually indicates that there are more of one sex in the population. We use a test called a called a chi² test (chi square; for more information on this and other statistical tests, see our Basic Stats section). This test assumes that males and females are equally easy to capture. In the graphs in figure 5, we have put a star over the bars that show significantly biased sex ratios, i.e. the sex ratio is significantly different from one to one.

Question:

• Does monarch sex ratio vary throughout the breeding season and during the migration?

Methods:

We captured butterflies at different times during the summer and fall in Minnesota, Wisconsin and Texas, and recorded their sex and other physical characteristics. All of the June - early August samples were collected during several research trips to our study site in western Wisconsin in 1997. From observations of wild monarchs in both places, we are quite sure that our late August and October 1998 samples represent some of the first monarchs to leave Minnesota and arrive in Texas, respectively.

Results:

Sex ratios of butterflies collected in the United States during the summer and fall are shown in Figure 5.

In June, the sex ratio was male-biased, although this bias was only marginally statistically significant (p = 0.059). In July, we captured significantly more females than males (p < 0.01), while the sex ratio was approximately even in August. Both of the samples collected during the migration (September and October) were male-biased.

Discussion:

There are two possible explanations for significantly biased sex ratios: the population from which we are sampling actually has a biased sex ratio, or the population has a one-to-one sex ratio, but there is a catchability bias (i.e. one sex is more likely to be captured for reasons other than simply being more abundant). Here we discuss both of these possibilities.

Biased population sex ratio: Most of the monarchs that we see in Minnesota and Wisconsin in June have migrated from further south. A possible explanation for the male bias in June is that males migrate before females, although further study would be needed to determine if this is really the case. It is also possible that males and females migrate at the same time, but males are simply more likely to be found in habitats like our study site. This site has a high density of milkweed, and other researchers (e.g. Zalucki 1993) have found male-biased sex ratios in milkweed patches. Zalucki hypothesizes that females leave milkweed patches after mating in order to avoid frequent encounters with males. However, the female bias in July, when milkweed density is the same, would argue against this.

The male-biased sex ratios in MN in September and TX in October may result from an earlier departure of males during the migration. Data on diapause (Liz Goehring’s study) suggest that males go into diapause sooner than females, and may thus leave for Mexico before females.

Catchability bias: In July, our site has many monarchs. We captured significantly more females than males, but are not convinced that this sample accurately represents the actual sex ratio. Some of our catching team is better at capturing females during the breeding season. Females tend to fly more slowly as they look for plants on which to lay their eggs, and stop for several seconds when they do lay an egg. Males tend to fly fast as they search for females. When there are few butterflies around, we tend to catch every monarch that we see, but when several monarchs are visible at once, we may have an unconscious bias to chase those that are easy to catch!

We do not think that a catchability bias is responsible for the biased sex ratios in June, late August or October. In June, we saw few monarchs, so were unlikely to bias our chases toward those that were easy to catch. Most of the late August samples were non-reproductive monarchs captured nectaring. In October, butterflies were also non-reproductive, and were captured at migratory roosts or nectaring. In both the late August and October samples, males and females were behaving similarly when we captured them.

Discussion of sex ratio studies