Okay, I've gotta test the hypothesis that having multiple hunters helps minimize the amount of time that hunters must spend killing and processing the animal -- in short, the idea is that more hunters equals less time killing and packing it up.

The tricky part is that there are multiple species and none of them have a whole lots of samples. For instance, there are only 8 squirrels in all, 3 killed by one hunter, 4 by two hunter, and the last by a group of three hunters. There are similar distributions for all 12 species, only one of which has more than 25 kills overall.

While I could figure out how to test the idea for one species, I have no idea how to test the more general hypothesis that more hunters equals less time.

Any thoughts or comments would be much appreciated.

You may want to try a regression rather than ANOVA, since both of your variables are numerical (ANOVA will work also, but ANOVA often works better with designed studies rather than observational studies).

if there were only one prey type. I think what's getting me is how to look at the general question of solo and accompanied pursuits. What makes it challenging is that every species is killed in a slightly different way so that pursuit times between species are not comparable. Killing squirrels is always going to be quicker than killing wild boars, for example, no matter how many people are helping with each. Setting up a simple linear regression model for one species is easy enough, but there are not enough data for any one species to reach any definitive conclusions. Between the seven species in the sample, there are about 70 kills in all, roughly divided evenly between 1, 2, and 3 hunters.

I guess I wondered if an ANOVA would be the best way to go about it in part because I've heard about paired ANOVA tests and that sounded promising even though I wasn't sure how to match up all the data (I've got more squirrel kills with 2 people than I do with 1 and 3 hunters combined).