Local- and landscape-scale variables shape insect diversity in an urban biodiversity hot spot

Abstract

Local community structure is shaped by processes acting at local and landscape scales. The relative importance of drivers operating across different spatial scales is difficult to test without observations across regional or latitudinal gradients. Cities exhibit strong but predictable environmental gradients overlaying a mosaic of highly variable but repeated habitat types within a constrained area. Thus, cities present a unique opportunity to explore how both local and landscape factors influence local biotic communities. We used insect communities to examine the interactions among local environmental variables (such as temperature and relative humidity), local habitat characteristics (such as plant community composition), and broad-scale patterns of urbanization (including biophysical, human-built, and socioeconomic variables) on local insect abundance, species richness, and species composition in Los Angeles, a hot, dry, near-desert city. After accounting for seasonal trends, insect species richness and abundance were highest in drier and hotter sites, but the magnitude of local environmental effects varied with the degree of urbanization. In contrast, insect species composition was best predicted by broad-scale urbanization trends, with the more native communities occurring in less urbanized sites and more cosmopolitan insects occurring in highly urbanized sites. However, insect species richness and abundance were >30% higher and insect composition was similar across sites that hosted either native or drought-tolerant plants, regardless of the degree of urbanization. These results demonstrate that urban insect biodiversity is a product of interacting mechanisms working at both local and landscape scales. However, local-scale changes to urban habitats, such as cultivating plants that are adapted to the natural environment nearest the city, can positively impact urban biodiversity regardless of location.

Keywords: Diptera; Hymenoptera; Lepidoptera; Los Angeles; citizen science; community ecology.

Figure 1

Raw (A) insect species richness and (B) insect abundance per trap day across all 12 months of the study. Values are mean ±SE. Values are mean ± SE. Different letters indicate significant difference in means (P≤0.05).

Figure 2

Raw (A) insect species richness and (B) insect abundance per trap day in sites in which drought‐tolerant plants were present across all 12 months of the study. Values are mean ± SE. Different letters indicate significant difference in means (P≤ 0.05). This pattern also was similar across the individual

Figure 3

Raw insect abundance per trap day in sites with and without mulch. Only collections made during June through August are included.Values are mean ± SE. Different letters indicate significant difference in means (P ≤.05).

Figure 4

Insect species richness per trap day over daily mean relative humidity (RH) readings measured at highly urbanized sites (urban type 8) for collections made from September through February.

Figure 5

Raw insect species richness per trap day across the seven represented urban types for collections made from June through August. Values are mean ± SE. Different letters indicate significant difference in means (P ≤.05).

Figure 6

An ordination of species composition throughout all surveys along nonmetric mutidimensional scaling (NMDS) axis 1 and 2 (stress =0.16). These axes were chosen to best visualize differences among seasons and along environmental vectors. Points represent individual collections. Sites are colored by season. A legend matching colors with season is provided in the top right of the figure. Ellipses indicate the standard error measurements around the centroid of each season. Vectors indicate significant correlations between species composition and a measured environmental variable. The length of each vector is proportional to the strength of the correlation.