Figure 3. Heatmaps and scatter plots showing the horizontal and vertical distributions of (A) air temperature, (B) relative humidity, (C) photosynthetic photon flux density (PPFD), (D) total leaf area, (E) ant species abundance (ln (x+1) transformed)), and (F) ant species richness. Colors with higher saturation indicate higher values of the variables. Plots with missing ant samples were highlighted by grey color in (E) and (F).
Distribution of ant assemblages along vertical and horizontal gradients
In total, 35,710 individual ants from 138 species in 31 genera were sampled. Ant abundance decreased significantly with increasing height in the canopy (ln (x+1) tranformed, R2 = 0.67, F 7, 59 = 16.68, P < 0.001) from an average of 395 individuals per m2 at 5 m, to 101 individuals per m2 at 60 m. Ant species richness also significantly decreased with height in the canopy (R2 = 0.66, F 7, 59 = 16.65, P < 0.001), from an average of 12 species per m2 at 5 m to an average of 3 species per m2 at 60 m. Both ant abundance and richness also significantly varied across transects (horizontal positions) (Table S3, Fig. 3). Ant community assemblages showed significant differences in species composition both vertically and horizontally ( (Fig. S1 and S2; PERMANOVA analysis, vertical stratification: F = 1.8, P = 0.01, R2 = 0.09,; across transects: F = 6.4, P = 0.001, R2 = 0.39,).
We pooled data within the same transect or vertical stratum to examine distance-decay patterns between transects/strata. This showed strong effects of vertical distance on pairwise dissimilarity indexes between vertical strata (MRM analyses: coefficient = 0.007, P = 0.001,R2 = 0.65, Fig. 4). In contrast, we found no effects of horizontal distance on pairwise assemblage dissimilarity between transects (MRM analyses: coefficient = 0.0003, P = 0.52,R2 = 0.02, Fig. 4). At distances of < 50 m vertically and horizontally, pairwise dissimilarity was consistently higher horizontally than vertically (Fig. 4).