Date of Award


Degree Type


Degree Name

Master of Science (MS)



First Advisor

Dr. James Watling


Aim We examined the scale equivalency of fragmentation effects at the patch scale and landscape scale to determine if patch level effects scale up to landscape level effects. First, we examined species responses to fragmentation effects at the patch scale and landscape scale. Second, we evaluated whether there is a difference in response between taxonomic groups and across latitudes. Finally, we analyzed the data at the level of individual species in order to determine how species respond to edge effects and fragmentation effects. Location Data for 71 studies were gathered from the BioFrag database. Studies used were located in 48 unique landscapes across the globe. Major taxa studied Our global dataset comprised 7619 species from 5 different taxonomic groups (1212 birds, 279 herps, 3490 invertebrates, 136 mammals, and 2502 plants). Methods We used vote counting in tandem with a meta-analysis comparing effect sizes of local-scale edge effects to landscape-scale fragmentation effects. We analyzed data at the species level using random placement models to determine how species individually respond to edge effects and fragmentation effects. Results Negative edge effects and fragmentation effects were not the most prevalent in our study landscapes. Nonsignificant responses to edges and fragmentation were more common than either negative or positive responses. Negative edge effects also do not scale up to negative fragmentation effects. Negative effects of fragmentation per se also do not predominate with species abundance permutations. Our abundance results show that species with significant negative responses were in the minority, indicating that individual species responses are not overwhelmingly negative. Main conclusions Overall, our results suggest that patch scale effects of fragmentation (edge effects) should not be extrapolated to landscape scale effects of fragmentation (fragmentation per se effects). In contrast to much of the literature, species responses were largely nonsignificant across both scales and individual species responses do not show a variation from this trend. Based on the results of our study, we suggest that researchers strongly consider the scale at which a study is conducted and recommend that researchers avoid untested extrapolation across scales.

Included in

Biology Commons