Phylogenetic signal of sub-arctic beetle communities

Abstract

Postglacial dispersal and colonization processes have shaped community patterns in sub-Arctic regions such as Churchill, Manitoba, and Canada. This study investigates evolutionary community structure within the beetle (Coleoptera) families of Churchill and tests whether biological traits have played a role in governing colonization patterns from refugial and southerly geographic regions. This study quantifies sub-Arctic beetle phylogenetic community structure for each family using the net relatedness index (NRI) and nearest taxon index (NTI), calculated using publicly available data from the Barcode of Life Data Systems (BOLD); compares patterns across families with different traits (habitat, diet) using standard statistical analysis (ANOVA) as well as phylogenetic generalized least squares (PGLS) using a family-level beetle phylogeny obtained from the literature; and compares community structure in Churchill with a region in southern Canada (Guelph, Ontario). These analyses were also repeated at a genus level. The dominant pattern detected in our study was that aquatic families were much better represented in Churchill compared to terrestrial families, when compared against richness sampled from across Canada and Alaska. Individually, most families showed significant phylogenetic clustering in Churchill, likely due to the strong environmental filtering present in Arctic environments. There was no significant difference in phylogenetic structure between Churchill and Guelph but with a trend toward stronger clustering in the North. Fungivores were significantly more overdispersed than other feeding modes, predators were significantly more clustered, and aquatic families showed significantly stronger clustering compared to terrestrial. This study contributes to our understanding of the traits and processes structuring insect biodiversity and macroecological trends in the sub-Arctic.

Keywords: Arctic; DNA barcoding; biogeography; entomology; environmental filtering; macroecology; phylogenetic community structure.

FIGURE 1

Phylogenetic trees demonstrating phylogenetic community structure patterns. Each habitat or geographic region is shown by a different colour and shape. (a) Pattern a shows a clustering pattern, where closely related species share the same region. (b) Pattern b shows an overdispersed pattern, where closely related species inhabit different regions or environments

FIGURE 2

Map showing the location of Churchill and Guelph and the sampling sites within Canada and Alaska for beetle data available from BOLD

FIGURE 3

Example phylogenetic tree with a chart showing nodal distances among members of the community. NRI uses all the distances to find the mean pairwise distance ((1 + 2 + 3 + 2 + 3 + 2)/6 = 2.16). NTI uses only the distances between nearest neighbors (nearest neighbor pairs: A&B, B&A, B&C, C&D; (1 + 1 + 2 + 2)/4 = 1.5)

FIGURE 4

(a) (i) Phylogenetic community metrics for Coleoptera families in Churchill, MB. (ii) Shows only the families also present in Guelph for ease of comparison. A positive value indicates a clustered pattern, and a negative value marks an overdispersed pattern. Families exhibiting significant (p‐value < .05) clustering are marked by an asterisk. The majority of families tend toward a clustering pattern. (b) Graph showing the clustering values for Coleoptera families in Guelph, ON. The phylogenetic community structure is generally random, without a clear trend toward overdispersion or clustering. Families are more overdispersed in this region than Churchill. (ii) Shows only the families also present in Churchill for ease of comparison. For all the graphs, families are organized by adult diet

FIGURE 5

Phylogenetic tree showing the terrestrial and aquatic families present in Churchill. This tree is based on the one shown in Zhang et al. (2018). The pie graphs show the percent of the total BINs from Canada and Alaska that have been found in Churchill. Aquatic families have a larger percent of their total BINs found in Churchill than terrestrial families

FIGURE 6

Boxplots showing the results of the PGLS for the clustering values of Coleoptera families inhabiting specific habitats using (a) NRI and (b) NTI, and the clustering values of Coleoptera families exhibiting different adult feeding modes using (c) NRI and (d) NTI, and larval feeding modes using (e) NRI and (f) NTI. The same letter above bars denotes groups that do not differ significantly, while different letters denote a significant difference; aquatic families are significantly clustered using the NRI, and fungivore families are significantly overdispersed using the NRI

FIGURE 7

Clustering values for Coleoptera families in Churchill, MB, after the sensitivity analysis. The results did not substantially differ from the original analysis, and the main conclusions were supported