Distance Dependent Contribution of Ants to Pollination but Not Defense in a Dioecious, Ambophilous Gymnosperm

Abstract

Dioecious plants are obligate outcrossers with separate male and female individuals, which can result in decreased seed set with increasing distance between the sexes. Wind pollination is a common correlate of dioecy, yet combined wind and insect pollination (ambophily) could be advantageous in compensating for decreased pollen flow to isolated females. Dioecious, ambophilous gymnosperms Ephedra (Gnetales) secrete pollination drops (PDs) in female cones that capture airborne pollen and attract ants that feed on them. Plant sugary secretions commonly reward ants in exchange for indirect plant defense against herbivores, and more rarely for pollination. We conducted field experiments to investigate whether ants are pollinators and/or plant defenders of South American Ephedra triandra, and whether their contribution to seed set and seed cone protection varies with distance between female and male plants. We quantified pollen flow in the wind and assessed the effectiveness of ants as pollinators by investigating their relative contribution to seed set, and their visitation rate in female plants at increasing distance from the nearest male. Ants accounted for most insect visits to female cones of E. triandra, where they consumed PDs, and pollen load was larger on bigger ants without reduction in pollen viability. While wind pollination was the main contributor to seed set overall, the relative contribution of ants was distance dependent. Ant contribution to seed set was not significant at shorter distances, yet at the farthest distance from the nearest male (23 m), where 20 times less pollen reached females, ants enhanced seed set by 30% compared to plants depending solely on wind pollination. We found no evidence that ants contribute to plant defense by preventing seed cone damage. Our results suggest that, despite their short-range movements, ants can offset pollen limitation in isolated females of wind-pollinated plants with separate sexes. We propose that ants enhance plant reproductive success via targeted delivery of airborne pollen, through frequent contact with ovule tips while consuming PDs. Our study constitutes the first experimental quantification of distance-dependent contribution of ants to pollination and provides a working hypothesis for ambophily in other dioecious plants lacking pollinator reward in male plants.

Keywords: Ephedra; Gnetales; ant pollination; dioecy; gymnosperm pollination; pollination drop; protective mutualism; wind pollination.

FIGURE 1

Reproductive structures and ant visitors of the dioecious gymnosperm Ephedra triandra (Gnetales). (A) Female cones, one with pollination drop (arrow); (B) cluster of male cones at anthesis; (C) Camponotus mus; and (D) C. punctulatus foraging on the micropylar tubes of the two ovules where pollination drops collect. Ants collected on female cones showing E. triandra pollen grains (white arrows): (E) C. blandus, with pollen on its head, mesosoma and legs, and (F) Forelius chalybaeus, with pollen on its head and antennae. Scale bars: (A,B) 5 mm; (C,D) 2 mm; (E,F) 0.5 mm.

FIGURE 2

The effect of pollination treatment and distance to the nearest male on seed set in Ephedra triandra. (A) Percentage of mature seeds relative to the number of initial ovules under three pollination treatments (N = 15 plants). Boxes show median (line), 25th and 75th percentiles, whiskers encompass the range of values, and dots represent outliers. Different letters indicate significant differences (post hoc Tukey’s test, P < 0.001). (B) The effect of distance to the nearest male on seed set for the different pollination treatments (N = 15 for each treatment). Lines show significant linear regressions with 95% confidence intervals (dashed lines) on log₁₀ (x + 1) transformed values. Significance: ***P < 0.0005, **P < 0.005. Note the log₁₀ scale used in both axes. Values are shown back-transformed for clarity.

FIGURE 3

Pollen load and pollen grain germinability after contact with ants foraging on the pollination drops of Ephedra triandra female plants. (A) Number of pollen grains per individual on four ant species (Brachymyrmex patagonicus n = 24, Camponotus blandus n = 30, C. mus n = 73 and Forelius chalybaeus n = 14 individuals). (B) Pollen germinability (percentage of pollen grains that grow a pollen tube) for three ant species carrying the highest pollen load, compared to Control pollen, collected directly from male cones without contacting ants (N = 10 per treatment). Boxes show median (line), 25th and 75th percentiles. Whiskers encompass the largest and smallest values, and dots show outliers. Different letters indicate significant differences in post hoc Tukey’s test.

FIGURE 4

The effect of distance to the nearest male on pollen flow by wind and on the rate of ant visitation to Ephedra triandra female plants. (A) Pollen abundance (number of pollen grains per slide) in pollen traps placed on female plants at increasing distance to the nearest male plant. Open pollen traps (black circles) and traps enclosed in mesh bags (red triangles), to control for mesh effects (n = 60 pollen traps per treatment). (B) Ant visitation rate to female cones (number of ant visits per cone in 30 min) at increasing distance from a male plant (n = 75). Lines show significant linear regressions with 95% confidence intervals (dashed lines) on log₁₀ (x + 1) transformed values. Significance: **P < 0.001, ***P < 0.0001. Note the log₁₀ scale used in both axes. Values are shown back-transformed for clarity.

FIGURE 5

The dynamics of pollination drop secretion in Ephedra triandra (under laboratory conditions). (A) Two pollination drops (PD) newly formed on micropylar tubes (mt) of the two ovules (ov) 12 h after manual removal, and (B) fused into one droplet after 24 h. (C–H) PD withdrawal in artificially pollinated cones. (H) PD completely withdrawn after 120 h:, residual PDs and pollen grains can be seen on the rim of the micropylar tube opening, and along its exterior. P = pollen grains. Scale bars: 0.5 mm.

FIGURE 6

Proposed model for the combined wind and ant pollination of dioecious Ephedra triandra. Wind is the primary pollen vector, carrying pollen grains (yellow circles) from male to female plants. When pollen lands on pollination drops (PD) secreted by ovules, fertilization is achieved. Ants become secondary pollen vectors when they redirect pollen on their bodies to the ovules while feeding on PDs. As the distance between female plants and pollen donors increases (+), airborne pollen flow decreases (–), more PDs are secreted (+), and more ants are recruited to collect the PDs (+).