Table 1.
Study site characteristics.
Table 2.
Qualitative network parameters.
Fig 1.
Floral abundance, during and after R. ponticum flowering
Total number of available floral units, comprised of invasive R. ponticum (light grey bars) and alternative plant species (dark grey bars), during the flowering of R. ponticum (R1) and after cessation of the flowering of the invasive species (R2) at four invaded Irish woodland sites.
Fig 2.
Non-metric Multi-Dimensional Scaling plot of floral abundance data
Each point on the graph represents a replicate of floral abundance sampling, n = 3 replicates per site per sampling round. The closer the points, the more similar the identity and abundance of flowering plant species. Light grey squares represent sites sampled during R. ponticum flowering and black circles represent the same sites after flowering of the invasive species ceased. Label codes indicate the sampling round (1 or 2), the site name (C = Crossover, D = Dunran, S = Shankhill, T = Trooperstown), and the replicate (1, 2, or 3). Data were square root transformed to balance contributions of rarer and dominant flowering species.
Fig 3.
Flower-visitor species richness and abundance data
Flower-visitor species richness and abundance during R. ponticum flowering (round 1, light grey bars) and after the cessation of flowering of the invasive (round 2, dark grey bars). Comparison between round 1 and 2 of a.) total insect, bumblebee, and hoverfly species richness and b.) number of visits of total insects, bumblebees, and hoverflies, c.) insect species richness to non-Rhododendron plants, and d.) number of visits to non-Rhododendron plants. Plots show mean values per round, averaged across the sites, and vertical bars show the standard error of each mean. Number of visits represents the number of floral units visited by insects, however insect abundance (number of individuals, not shown) followed the same patterns. Significant differences in the above parameters when comparing the two rounds of sampling are indicated by * (paired t-test, p < 0.05).
Fig 4.
Non-metric Multi-Dimensional Scaling plot of flower-visitor interaction frequencies
Flower-visitor composition at sites invaded by R. ponticum during (round 1) and after (round 2) flowering. Site level mean interaction frequencies of pollinator groups were used to calculate the resemblance matrix. The closer the points, the more similar the identity and abundance of interaction frequencies of flower-visitors recorded. Data were square root transformed to balance contributions of rarer and dominant insect visitors.
Fig 5.
Quantitative insect-flower interaction networks
Networks represent insect-flower communities at four Irish woodland sites during (a, c, e, g) and after (b, d, f, h) R. ponticum flowering. For each web, upper bar widths represent pollinator guild abundance while lower bar widths are determined by the interaction strength with insect species. On the lower bars, the number 1 corresponds to R. ponticum and is circled in gray; remaining species codes and pollinator guild abbreviations are listed in S1 & S2 Tables. Linkage width indicates the frequency of the interaction. One network was calculated for each site during each round of sampling.
Fig 6.
Quantitative network parameters
Comparison of quantitative network parameters calculated from insect-flower interaction networks created by sampling four invaded Irish woodland insect- communities during (light gray) and after (dark gray) flowering of Rhododendron ponticum. Network parameters analyzed include a.) quantitative connectance, b.) interaction evenness, c.) generality, d.) vulnerability and e.) H’2. * indicates a significant difference between the two rounds of sampling (paired t-test, p < 0.05). Plots show mean parameters per sampling round across all four sites, and vertical bars show the standard error of each mean.