How do networks vary among southern, middle, and northern ranges?

Here I assigned the sites to subspecies ranges based on visual inspection of the USDA plants database. I reduced the number of samples to the 21 most northern (7 sites) for comparison to the southern subspecies dataset. A middle range was selected by measuring the geographic distance between the southern and northern sites and choosing the seven sites closest to the middle. The top taxa at the OTU level are selected for computational efficiency.

Plots of the networks by geographic location and taxonomic level

Network characteristics

Betweenness centrality

Clustering Coefficient

Error bars indicate the mean and SD of Erdős–Rényi random graphs.

Average path length

Error bars indicate the mean and SD of Erdős–Rényi random graphs.

Modularity

Error bars indicate the mean and SD of Erdős–Rényi random graphs.

Mean degree

Testing if networks have scale-free property (follow a power law)

A significant p-value indicates that the network does NOT follow a power law.

## [1] "OTU south"
## [1] 0.1530834
## [1] "OTU middle"
## [1] 0.2725367
## [1] "OTU north"
## [1] 0.2888733
## [1] "Genus south"
## [1] 0.002623463
## [1] "Genus middle"
## [1] 0.8065674
## [1] "Genus north"
## [1] 0.3450296
## [1] "Family south"
## [1] 0.9999996
## [1] "Family middle"
## [1] 0.9742237
## [1] "Family north"
## [1] 0.5193837

How does the incidence of positive and negative co-occurences compare across latitude?

Histograms of the rho values in different regions

Test of difference between the distributions of rho values across latitude

Compare only significant co-occurrences across latitude