Minimal impacts of invasive Scaevola taccada on Scaevola plumieri via pollinator competition in Puerto Rico

Abstract

Introduction: Scaevola taccada and Scaevola plumieri co-occur on shorelines of the Caribbean. Scaevola taccada is introduced in this habitat and directly competes with native dune vegetation, including S. plumieri, a species listed as locally endangered and threatened in Caribbean locations. This study addresses whether the invasive S. taccada also impacts the native S. plumieri indirectly by competing for pollinators and represents the first comparative study of insect visitation between these species.

Methods: Insect visitation rates were measured at sites where species co-occur and where only the native occurs. Where species cooccur, insect visitors were captured, identified and analyzed for the pollen they carry. Pollen found on open-pollinated flowers was analyzed to assess pollen movement between the two species. We also compared floral nectar from each species by measuring volume, sugar content, and presence and proportions of amine group containing constituents (AGCCs).

Results: Our results demonstrate that both species share insect visitors providing the context for possible pollinator competition, yet significant differences in visitation frequency were not found. We found evidence of asymmetrical heterospecific pollen deposition in the native species, suggesting a possible reproductive impact. Insect visitation rates for the native were not significantly different between invaded and uninvaded sites, suggesting that the invasive S. taccada does not limit pollinator visits to S. plumieri. Comparisons of nectar rewards from the invasive and the native reveal similar volumes and sugar concentrations, but significant differences in some amine group containing constituents that may enhance pollinator attraction.

Conclusion: Our analysis finds no evidence for pollination competition and therefore S. taccada's main impacts on S. plumieri are through competitive displacement and possibly through reproductive impacts as a consequence of heterospecific pollen deposition.

Keywords: Goodeniaceae; Scaevola plumieri; Scaevola taccada; invasive; pollination.

Figure 1

Floral morphology of (A) Scaevola taccada and (B) Scaevola plumieri. The stylar indusium (a synapomorphy for Goodeniaceae) is located above the half-flower in Scaevola species. The indusium grows up through mature anthers and secondarily presents the pollen to pollinators, then later become a receptive stigmatic surface. At our study locations, we frequently observed flowers of S. taccada to have purple streaks not observed in S. plumieri; corolla lobes of S. plumieri appear fringed, a morphology not observed in S. taccada.

Figure 2

Puerto Rico sampling and observation sites. (A) Study locations (stars) on the main island include Camuy (18.489145, -66.845474), Piñones 1 (18.450074, -65.951980), and Piñones 2 (18.446056, -65.933833). (B) Study locations on smaller islands (island sizes not to scale) include the Vieques study site at Playa Grande (18.08847 -65.51550) and the Culebra collection sites at Playa Flamenco (18.328483 -65.316855) and Playa Brava (18.329939, -65.284600).

Figure 3

Comparison of overall insect visitation frequency for Scaevola plumieri (SP) and S. taccada (ST). Insect visitation frequencies are shown for S. plumieri (green boxplots) and S. taccada (blue boxplots) at the invaded site (Playa Grande) for 2015-2017 and for S. plumieri at the uninvaded site (Camuy, orange boxplot) and two invaded sites (P1 and P2; yellow boxplots) for 2020. Lines inside the boxplot represent medians, the bars above and below represent maximum and minimum values, and circles indicate outliers. Mann Whitney U tests revealed no significant differences between species in any particular year, nor between species all years combined, nor between S. plumieri at invaded sites (Playa Grande) vs. uninvaded sites (Camuy). Plot produced in R (R Core Team, 2023).

Figure 4

FE-SEM images showing a polar view of Scaevola plumieri (A) and Scaevola taccada (B) pollen grains. Light microscopy images are shown in the insets. Both species have tricolporate microspinulose grains (further characteristics of Scaevola pollen are described in Gustafsson et al., 1997). S. taccada pollen grains are slightly larger than S. plumieri, but otherwise appear similar.

Figure 5

Analysis of pollen size and pollen transfer between Scaevola species. (A) Frequency distribution of pollen diameters plotted as a density graph from unopened flowers of S. plumieri (peach) and S. taccada (teal). (B) Comparison of pollen sizes from unopened flowers of seven individuals of S. plumieri (peach) and nine individuals of S. taccada (teal). ANOVA test shows significant differences in pollen sizes from individuals compared to the overall mean (dotted line). ****p ≤ 0.0001; **p ≤ 0.01; *p ≤ 0.05; ns = not significantly different. Overall difference (ANOVA) between species is significant (p < 2.2 x10^-16) (C) Identity of pollen grains found on open-pollinated flowers of each species per year, expressed as a percentage of total pollen grains assigned. Based on pollen size distributions shown in (A, B) pollen grains > 37.73 µm were assigned to S. taccada (ST; teal) and pollen grains < 32.95 µm were assigned to S. plumieri (SP; peach). Plots produced in R (R Core Team, 2023).

Figure 6

Interspecific variation in composition of amine group containing constituents (AGCCs). Mini-plots compare proportions of 16 different AGCCs between S. plumieri (peach) and S. taccada (teal). Four constituents (AGCC 2, 12, 14 and threonine) had significantly different proportions in the total AGCC content between the two species. S. taccada has proportionally higher amounts of unidentified AGCCs 2 and 14 whereas threonine and AGCC 12 are proportionally higher in S. plumieri. Constituent 2 was not detected in S. plumieri. Norleucine was used as the internal standard and could not be distinguished from native norleucine and is therefore not included here. Y-axis values represent proportions of each constituent to the total AGCCs present (mass of AGCC/total mass of all ACGGs).