Supporting information for Lockwood (2003) Proc. Natl. Acad. Sci. USA, 10.1073/pnas.0535132100.

Table 4. Results of testing for abundance differences between victims and survivors, controlling for size, across all localities at the subgeneric level

Spatial scale

Abundance metric

Raw

Rank

Proportional

Overall (cumulative)

F

1,90 = 1.88, p = 0.17

F

1,90 = 1.56, p = 0.21

F

1,90 = 1.88, p = 0.17

Overall

(averaged)

F

1,90 = 2.18, p = 0.14

F

1,90 = 0.23, p = 0.63

F

1,90 = 2.58, p = 0.11

Controlling for size did not affect the relationship between abundance and survivorship. Three abundance metrics (raw, rank, and proportional) and two data treatments (cumulative and averaged) were analyzed in an analysis of covariance, with size as a covariate. No significant differences were found between victims and survivors regardless of abundance metric used or treatment of data.